Building Things to Help you Sell the Things you Build | Patrick McKenzie | BoS USA 2013

There are many software companies that are world class at building products. Great products. There are a lot fewer software companies that are world class at selling those products. Patrick McKenzie has some observations on why that might be and at BoS2013 he proposed some solutions that any company could put in place. Put your hand up if you are convinced of the importance of any of the following for your business:

• SEO/Adwords
• A/B testing
• Conversion optimisation
• Funnel analytics
• Onboarding/first-run experience
• Lifecycle emails/drip marketing

Now. Keep your hands up if you think you’ve totally got these sorted. No room for improvement. Hmmm, I’m not seeing a whole load of hands out there. You need to watch this: 45 minutes of insight from the ever excellent Patrick McKenzie on how to embed marketing development in your business as a repeatable practice. Why it isn’t easy, how you can get started and the single most important thing you can do to improve your growth. Patrick finishes with a moving personal testimony on the importance of discussing mental health issues in the tech industry. Since we filmed this and Greg Baugues’ amazing talk, we have had a huge response, thanks to everyone who has come forward to speak about their situation. You are all making a difference.

Video

http://businessofsoftware.wistia.com/medias/og2a7kswfk?embedType=async&videoFoam=true&videoWidth=640

Transcript

Hi, hello everybody my name is Patrick McKenzie. Perhaps better known as Patio11 on the internet. So I was saying that I was a little worried about this presentation and somebody said you’re going to do a fish out of water story from Japan, everyone’s going to get a laugh out of it and then you’re gonna do the presentation like you always do. And I’m like darn, they know my trick. [laughter]

So I’m gonna do something totally different this year. I’m gonna tell you a fish out of water story about Brazilians but living in Japan. So there’s about a thousand Brazilian people in my town and according to Japanese people they all look exactly like me. And uh…[laughter] So recently I was going to city hall to pay my taxes and I go in there…city hall is about 1,000 people who work at it. It’s 15 different divisions so they have a receptionist who is strategically placed directly on the, inside the front door to tell you, you need to go to the first floor, turn left at the water department to go to where you’re going. Yadda, yadda. And she sees me and says good morning in Portuguese. And I don’t actually speak Portuguese but having been a consultant for a few years I know how to repeat things without understanding them, so I say good morning right back at her. [laughter] And then she gets that look on her face -I’m sure you’ve all seen this one before- the I have something complicated to say but I don’t speak enough Portuguese so I’ll try saying it in Japanese and just pray you understand look. Hmm? And says: The welfare department is on the second floor. I’m like, okay…I’m actually here for the revenue department. And uh…And she says why, why on earth would you be here for the revenue department? And I say well I guess I pay for the revenue…I pay for the welfare department. [laughter]

So business in South Korea is always good news for me. The first time I came to visit to South Korea to talk I had met the woman who would become my wife right before leaving for the trip. We went on our first date probably a week after I got back from BoS. The second time I spoke at Business of Software I proposed to her and she said yeah…oh shoot, ha-ha. Forgot which one was the next slide. So if you want these slides they’re up at the Bitly thing. I suggest downloading them later rather than right now. Yeah, BoS a nice thing. She said yes, so this is our photo from our wedding last year. [applause] I get a lot of mileage out of this photo because I’m never gonna look that good again.

So what do I do business wise? I run Kalzumeus Software. It’s a really small software business. My original product was Bingo Card Creator. It makes bingo cards for elementary school teachers. I transitioned to Appointment Reminder, it does appointment reminding phones, SMS messages and emails for professional services businesses. And I also do a bit of a training business teaching other software companies on how to sell more software. Until earlier this year I did consulting in that too. Transitioned out of consulting. Ask me why later and you’ll probably need to get me drunk to get the full story which is problematic because I don’t drink. [laughter]

I always kind of feel like a fish out of water at the BOS conference. I remember the first year I was here I turned to my left and said what do you do? And he said I run a 20 man software company, we build a data visualization and control software for nuclear power plants and had revenues of $50 million last year. What do you do? Bingo cards for elementary school teachers…[laughter] So you know, don’t put me in the league of the HubSpots of the world but I have a fairly fun little business and have learned a few things doing it and I’d like to share some of them with you.

I’ve also…so in the course of building up this little software business I got good at a bunch of things that are not classically development but involve actually programming. So SEO, AdWords, email marketing, yadda, yadda, yadda. And the weird thing was these things that worked for selling bingo card to elementary school teachers turned out to work on other products too. And Charles Polski gave me some advice one day, he said you know, the skills that you have built up on selling that bingo creator would probably be more lucrative if it were applied to a product that wasn’t totally bullshit. Uh so…[laughter] so I started consulting because there are many software companies in the world that are pretty good on the building products. Which I have no special talent for. And often software companies being built by software people don’t really understand the marketing stuff all that well. And I was able to explain a couple forms of marketing in way that they would accept and got some fairly decent results for them.

I don’t want to talk so much about the awesome successes early on here. We’re gonna talk a little bit about the failures because like, had some successes with the software companies but we you know, and my fault here cause I was the consultant trying to get this implemented. But we did not achieve like long term lasting success in a lot consulting engagements. Like we achieved over the course of the engagement but did not successfully effect a change in the way that company did business that lasted years and years down the road. And I think, can I just have a show of hands here? Who here is pretty sold on the wisdom of doing these kind of things in the sof- in their company? Like hands up, great. So I won’t sell you again on doing the wisdom of these things but who here is like, we’ve totally got this stuff nailed we have nowhere to improve on it we’ll go back to product. No hands right?

So here are some ways that you might be having some difficulties in achieving this as a repeatable practice in your company.

And this the most important takeaway from both this talk and my consulting career. Which I never, one of the reasons I didn’t successfully sell this to customers as something they should be doing is that I never put it in a single sentence. And so it’s like you should do A/B testing and email marketing and all these little tactical things without saying the overarching strategic picture. The overarching strategic picture is that you should definitely have full time engineers whose only job is achieving marketing and sales objectives. They don’t have to work on product, they don’t have to work on SQL optimization, they don’t have to make sure the server is running at night. The only thing they have to do is whatever the KPIs you give them, whether that’s conversion rate, turn rate, whatever. Like that is the number they have to hit. And code that is written towards that end is a win for them. Code that is written towards any other end is something they shouldn’t be messing around with.

So why aren’t we all doing that already?

There’s typically four main failure modes that I see working with consulting clients and in my own businesses. Because you’re too busy in the day-to-day of the business as an engineer. Because the company has structural or organizational or cultural issues where we know we want to do this but it will cause an absolute knife fight between the engineering team and the marketing team to actually get that accomplished and so it doesn’t happen. Because we thought we could just like buy some SaaS for $29.99 a month that would take care of this and that didn’t work out so well. And we want to do it, we know we should do it, we have the internal will to do it but we can’t find someone internally who’s got the skill set for doing this.

So some ways you can possibly overcome these. Or maybe dig into the pathology a little bit more first. So I’m an engineer and one thing I love about actual development, it gives you a great definition of progress. Progress is measured in units of working code. Which it works great if you’re doing actual development in the context of a bank or something because you have a clear business model that is succeeding. Like the bank borrows money from people, lends it to other people. There’s a dealt in the interest rates charged and as long as the bank successfully executes on that we know they make money. So the software just needs to kind of just grease the wheels of the successful execution of the core business model. Unfortunately in the software company, software doesn’t just have to grease the wheels, it is the core business model. So I think progress in a software company is not measured by just having code that works but it’s measured by…Hate sounding a little bit mercenary but just top flight revenue growth is really a good metric to focus on. And often as an engineer we get boiled down in the weeds of focusing on things that feel to the engineer in us that they’re really important but don’t move that tight line revenue growth needle at all.

I have a really tragic story and this is…we never call a consulting client by names, this is distilled from a couple stories and maybe fictionalized a little bit but not all that much. We had a brilliant engineer who was working with me on a project at a consulting client and they got into A/B Testing with a will and a way they had a genuine talented interest for it. And in the course of two weeks of implementing A/B test they increased that company’s enterprise value by 5%. Yeah. Whoa right? And so I came back a year later and I said okay, your first two weeks on this effort you increased the company’s enterprise value by 5% so my expectation right now is that the company has had the best year ever and that you are currently being compensated like, you know, more riches than God, right?

And he said well actually, no I’m still doing pretty much what I was doing and actually today I’m kinda busy optimizing SQL queries on a report generation screen that’s seen by 0.5% of our customer base cause it’s slow. I said who made that decision? And he said well I’m the best SQL optimizer at the company and the screen was slow so I had to optimize it. I’m like how ’bout we hire like 20 SQL optimizers and just have you do the A/B testing cause you’re really good at that? And I took that up the chain of command to a couple places to that company and they just…engineering minds that didn’t quite gel with that way of looking at the world…and I think that’s something perhaps not as dramatic that happens in a lot of companies. Different issue that some companies have.

So often in my consulting career I got called to be a marriage counselor between the marketing team and the engineering team. And just like marriage counselors are supposed to be the neutral party that can kind of hear both of their things and get them to rephrase them to the other party in ways that say there’s unconditional love, support and trust. But these little niggling issues. Like the engineering team is like I don’t like when you make me add these little javascript snippets to the website and then you spend $20 million on that total bullshit. And then they’re like I respect that opinion. I don’t like when you call my core competence total bullshit and when you…[laughter] And I also don’t like when you tell me that taking- adding a snippet of javascript to a website will take three weeks and a written change order. [laughter]

And yeah…A lot of software companies are set up with defined marketing and engineering teams just because that’s the way software companies have always been done. I think that long term that’s very problematic especially with the more agile, like small “a” agile software being done in the world today. But even if your company is currently structured like that you don’t necessarily need a full line restructuring to capture some of the benefits. We’ll talk about that later.

So I don’t know what you’d call an engineer that does marketing objectives like Devops seems to be a thing so if I just smash parts of words together, maybe Markdev or Devmark or whatever. But I’m told that they are absolutely impossible to hire. It’s like finding a freaking unicorn in the words of one consulting client. I think there’s ways to find them. We’ll talk about that in a moment.

So much love for Optimizing, Visual Website Optimizer, Kissmetrics. There are lots of wonderful tools to do a lot of these tasks that didn’t exist a couple of years ago. Unfortunately I have some clients who are like we totally signed up for a Visual Website Optimizer account two years ago. I’m like great, how many A/B tests have you implemented in the last two years? Four? Oh.

And not to pick on my consulting clients because I have a fairly good idea what the following show of hands is going to show. Raise your hands if you have ever done A/B Testing. See this is what I would love to see at this conference. Keep your hands up. All right, now put your hands down if you implemented a A/B test last week. And all the hands stay up. Or most of the hands stay up. A lot of the people know what they should be doing but don’t actually make a practice of doing that in the day-to-day basis in their companies.

Okay so how do we get an engineering team onboard with alright we’re going to have some of you guys focus your full time efforts not on products, not on building features, not on these fun engineering challenges you wanna solve but on these – integrating API with our mail service provider so that we can do life cycle emails better? One way is exploiting the hacker ethic that a lot of engineers have. It’s satisfying to win games if you’re a hacker but it’s even more satisfying to win by cheating or win by a case that the rules did not envision.

Funny story about that. So we were playing a game in one of my few memories of high school, there was this pseudo investment game done for one of our World History courses and teacher said the ten teams in the room are in competition with each other. Pick which industry you’re gonna invest in for each of the twelve periods of this game and then we’ll figure out who wins at the end. And I said how do we determine who wins at the end? And they said there is a point scale which if you get one point in industry X, it’s worth X amount of points. You get two investments in industry Y, it’s worth Z amount of points. Yadda-yadda.

So I looked at the point scale and I said the absolute optimum strategy is to buy railroads under every possible circumstances. We should buy railroads under every possible circumstances. And my team was like we should diversify. I’m like, in the real world that would work, but in this game, railroads. It is railroads, railroads and railroads. And so they overruled me and so we had a diversified portfolio and then the shadow portfolio that I was maintaining in real time of all railroads because I knew it was gonna win. And sure enough at the end of the game the shadow portfolio did win and the teacher said you probably should have listened to that. So if you can tell the engineers like hey, psst, don’t tell anybody else in the company this but I bet you could add more to sales by doing this weird little trick for two days than they have done for the last two product releases which took six man years of investment. Like, that’s totally crack to the engineer mindset.

Another way…So you have to over communicate on results. So I think humans are very status seeking animals and in our little, you know, we seek for it in our tribes and in the tribe of engineers like you know, getting your ideas, your implementations into the visible product is what generates status. You have that new awesome widget you did. You have the report running everyday, you have that awesome API integration that 20% of the user base is using and that’s awesome and people will get to see it and touch it and you get to like, mind ether and come up out things that are made out of pure dreams. And that’s why we do engineering. Or you could like twiddle some dowels for the marketing team. And mine stuff of pure dreams, twiddle dials for the marketing team’s bullshit.

So people do the fun stuff. So you need to communicate that they are like active ongoing experiments even if they’re not the see it, touch it experiments that you would get with interacting with the website and communicates those experiments are having real value. One of the most successful companies that I saw doing this they just had…the guy who was doing this it was a routine thing, every week he would say okay, so Monday here was my hypothesis, he was the experiment I started. Friday boom we stopped the experiment, here’s the results and then let me do some multiplication. Here’s how much money I made the company this week. And there was just an unceasing drum beat of these Friday reports. Every week for a year. His teammates knew he was killing it and not just wasting time on copy-pasting javascripts snippets for the marketing team. His superiors knew he was killing it. He knew he was killing it.

Because a lot of engineers we have that inferiority complex on our shoulder and if you hear users tweeting at you, oh that new thing is awesome! You feel I built that new thing, I am awesome! Because people very rarely tweet about dude that A/B Test that you put me through, that totally increased my propensity to convert by 3% making you $6 million this year. Awesome you! So that kind of like gives the synthetic feel of achievement.

To do this you’re probably gonna need to surface business metrics more closely to the engineering team. Again hacker mindset. If you give us a number, we have to make the number bigger and if you give us a progress bar, we have to move the progress bar to the right. We just have a genetic mutation that makes us absolutely incapable to resist this. So give us the right metric. Like top line revenue, great metric and say all right with the asterisks please don’t do anything unethical, immoral or would pass muster at Zynga make this number go higher. And it’s obligatory that I get at least one dig in Zynga in any presentation.

So humans are not just status seeking animals because we love status, there’s definite career wins to being seen as being successful in your organization. It gets you the better assignments within the organization, more actual benefits and makes life easier for you when you transition to your next job. Which given the realities of our industry we all know we’re going to be probably in a very different thing three years from now, five years from now than what we’re doing right now. So you need to say that look, this is not like the terrible scut work of the business. Copy-pasting a javascripts snippets. It’s not like you know, being build guy. Is anybody in here the build guy?

Audience Member: Yes! [laughter]

Patrick McKenzie: See it’s amazing how often the build guy is the CEO of a well run software company and the reason…[laughter] …the reason that the build guy is the CEO is the CEO knows that nobody else in the company is gonna do that God awful stupid work so I’m gonna have to man up and take one for the team. You don’t want people to think like oh, raising our revenue by 20% this quarter is just God awful stupid work so I don’t want to take one for the team. I’ll let someone else do that. That’s for sort of like a core thing for the business right? So you wanna say look if you’re successful at doing this you’re going to get internal recognition, you’re going to get the perks commensurate with somebody who just increased revenue by 20% and when you go on your career path both within this company and if you decide to move on in your adventure it’s going to be up and to the right.

So how do we change the culture and of the company that might be a little bit of a dysfunctional family unit so that we can know and appreciate each other? One of the things I love doing with starting with companies who aren’t really doing this as a practice is to focus on low hanging wins first. So there’s like a spectrum of implementation difficulty from let’s design a totally new onboarding experience. Which is going to take us six weeks of work from three team members and require a new release of the software.  It’s a lot of work. Versus why don’t we copy-paste in the Optimizely A/B Testing script and then run one A/B Test on the H1 of the homepage which is like 15 minutes of work right? The H1 of the homepage, just for people who love A/B Testing is one of the highest return per unit of time investment opportunities anywhere on a website.

So if you just run a few tests on that, all A/B test you typically get about three non results. No significant change to the business for everyone win but you four or five tests and you get your report back to people. Okay for a total time investment of two hours we just increased the number of trials we got by 10%. Which is equivalent to bumping our marketing spend up by $500,000 a quarter. So that was a win. And that sort of thing generates positive momentum internally and it kind of like gets over objections because the engineers who might have a very data driven mindset won’t argue with the data so much, they would argue with somebody on the marketing team who said hey I read this really good advice the [00:21:02] .com blog. I know you don’t really follow that or actually trust them but trust me they’re like good stuff. And the engineer is hearing blah, blah, blah, blah, blah, blah. Where’s my SQL?

Get positive momentum early. Again hacking with the presumed culture of your engineering team, show them they’re fun, interesting challenges involved outside of the product. So total confession time, neither of my two software products, Bingo Card Creator or Appointment Reminder is very interesting technically which means that as engineer I should be bored out of my skull. The only thing that keeps me interested in those businesses is the business around the product is like really fascinating. I can really sink my teeth into the A/B Testing and marketing stuff. So you can teach people that you can really sink your teeth into that stuff just like you would sink your teeth into getting a really exceptional user experience, or learning a new javascript framework, or dealing with Ruby on Rails security vulnerability releases, or creating them. This will help you keep your team more focused on that.

On the marketing side one the issues I have with proposing this to marketing teams…So I’m the ambassador from the marketing team to the engineering team I’m also the ambassador from the engineering team to the marketing team. I’m like, okay, so I just sold the engineers on doing a whole lot of data driven stuff to improve the marketing of the company. And what marketers hear is okay you’re gonna replace me with like a cyborg. It’s going to be like Terminator and Judgement Day is going to be I lose my paycheck sufficiently sophisticated A/B testing algorithm.

I say no, no that’s not the purpose. We’re just going to have the engineering team willingly, happily provide you tools which will make you even better at that awesome secret sauce that you do so well. While meaning you spend less of your time on the stupid scut work that you’re doing like manual report writing. And less of your time on kind of like throw a bunch of stuff at the wall, see what sticks when you could have a computer do that for you in 1/10th of the time at 1/10th the cost.

This is the hippiest I will ever sound in any presentation but I really sincerely mean this:

emphasize mutual respect, empowerment and opportunities to learn among all members of the organization

If people actually say the bullshit line in a meeting, lock that down. Look we are all on the same team here. We all have singular objective. We are making wonderful software for wonderful users which will improve their lives. And we’re all on…that is all of our overarching goals. The marketing team, the engineering team, the sales team we are all essential parts of that. We have a culture of mutual respect and support within this organization. I think most of you are like that. There are some companies in the world that are not. If you ever transfer into one, transfer out ASAP. Or change the culture.

Okay, so tools I mentioned some. There’s lots and lots of marketing/engineering tools that are popping up these days. Largely as less technical companies figure out I have no engineers on staff so I have no option for this whatsoever unless I buy it prepackaged in a box. So if you’re an insurance company that doesn’t have anybody with deep engineering skills, building your own A/B testing framework in a week just can’t happen. So buying Optimizely, buying Visual Website Optimizer is a great option for you. And then you can hand it to your nontechnical marketing team and say go nuts testing value propositions. Like your kids won’t starve to death if you die versus your wife won’t necessarily need to remarry if you die and see which has higher conversion rates.

A lot of people they have the tool, they think having the tool substitutes for doing the work. Very rarely does. One of the keys of getting the most use out of your tool, typically all the tool makers have really good blogs telling you like…I’m sure if you go to the Optimizely blog they’ll have a long history of case studies of people doing excellent things with A/B testing. Cause obviously it’s in their business to tell you how to do this better. That makes it more likely that you successfully do A/B testing and continue subscribing to them.

So learn what the community around your tool is doing with it. That works really well. This is especially true with tools that are kind of in emerging spaces like email marketing for software companies is very emerging space. So I would bend the ear of your email marketing company. They ask for best practices. Also learn…Share what you do internally too. So a best practice is having…pick whatever forum factor you like. Wikis work for a lot of people. Like just internal Wiki with here’s a list of all the A/B tests we did. Here’s why we ran it. Here’s what we learned and stars for things that are like, this was a huge amount of money for us. Or this wasn’t a huge amount of money but it was sure a surprise. Like nobody expected the test to turn out that way.

A/B testing you will frequently find that everything you know is wrong. Let’s see if I have a funny story of an A/B test. Okay one from my own product. I’d been working on Bingo Card Creator for six years at one point and then I showed it to a bunch of Ruby on Rails coders in the course of the demo just to show them how to do A/B testing in Ruby on Rails. I said hey, anybody in the room, you’re not gonna do better than my H1 title that’s on the first page because you haven’t been in this business for six years but just shout out something and I’ll just toss it in here in an A/B test and we’ll see who wins. And yeah, they crushed me by 10%. With less than 45 seconds of exposure to my business. So, shows what I know. But not an impediment to being a successful consultant on it. [laughter]

Patrick: Anyhow, also so I think just using an off the shelf tool is often a wonderful solution for less technical companies but you get so much more from deep integration into your own systems. I can’t really talk specifics about these things. Because it varies on a company to company basis. But people who inside their application have a really good understand of what users are doing. Tight integration of that understanding with your email marketing for example or with your A/B testing. Or tight integration of your particular mix of customers coming in via different channels with say your landing pages has extreme amounts of leverage built in. It means you do a little bit of work and you get a huge result out of the back end. So I would encourage you that you should probably be a comparative power user of these tools. Like obviously using just the basic feature set is better than not using it at all but after you’re doing the basic feature set you should be thinking okay they have an API for a reason, my company is probably the best situated of all the companies to use the API so I should probably be doing that.

So if you’re trying to hire unicorns, how do you get them on your team?

The easiest way and that works best for a lot of companies is go to your engineers that are a little bit open minded and kind of train them on the marketing stuff. Theoretically I suppose you can go to your marketers and train them on coding I just find that getting someone to the point where they’re comfortable solo pushing code into a production environment from a marketing background is often…that takes awhile. That’s a six month, one year project for them.

Whereas teaching an engineer about A/B testing and saying look you can’t possibly hurt the business with this because we’ll have a number that says if you are an we’ll just push the stop button if it doesn’t work out. So think of one sentence that you would replace the title tag with. Throw it up there the next time you edit that page and we’ll just see what happens. And something that somebody can learn to do in an hour. They’ll get better at it over time obviously but I’ve seen people who go from having no pre-existing marketing background to being huge force multipliers for their companies in literally a period of just weeks. My typical consulting engagement would be 2-3 weeks and the people who really hit the ground running would by the end of the three weeks be basically chargeable at many consultancies for that sort of thing.

Another option. So let’s say okay your engineers are just too indispensable or they just don’t want to do this. Maybe you can hire somebody. So, something you’ll hear pretty often in Silicon Valley or in the Y-Combinator circle of friends is that there’s a bunch of people that are in a particular cohort based on dates or whatever. Some of them their businesses they take of but they don’t quite achieve escape velocity and so they need a new gig. Now in the course of starting a new software business you often learn about the marketing stuff very very rapidly because you’re forced to or you die! That’s how I learned it. The software isn’t gonna sell itself so I guess I have to learn how to do that. So if you go to these folks who they achieved some knowledge of how to market and sell software but they didn’t have product-market fit or they achieve it fast enough to get their objectives. You might be able to hire them and say look, we know you know how to do…you’re a good enough engineer to work at this company. You’ve learned a little bit about AdWords or SEO or whatnot based on some blog post you wrote or what you’ve just told us in the interview. How about we hire you. Not to work on our product but pretend that you owned it and do this setup thing for us and your goal will be increasing revenue or increasing the number of trials we get per month. Or increasing our conversion rate. And give them some room to run on that. That works out pretty well and often these folks tend to be young and cheap. You don’t necessarily need somebody who’s senior vice president of engineering to have that experience.

There is the last option. I hesitate to mention it but I will for completeness. There are companies that are identified with doing this stuff really well. There’s a team in Facebook and a team in Google that specialize in basically this category of things. There’s a few other companies, Dropbox, that are publically identified with it. You could hire an alum from one of those companies and get them to do it for you. Problem with that, they are more expensive than almost anything you can imagine. Has anyone here ever hired a quant from Wall Street? Okay, just background information a quant from Wall Street is somebody who’s firm absolutely knows with absolutely religious certainty that his code individually made us $3 million last Thursday and as a result their pay packages are just absolutely gianormous. Like $250,000 base and then in a good year they might give you a bonus of several million dollars.

Now think of a company operating at the scale of Facebook and you have an engineer who’s core competency is increasing the number of people who click on those lovely little Zynga ads by like 2% and then do the multiplication there. And now if you get in a bidding war with Facebook over that engineer, you will not win. I had a company approach me and he said you seem to know this metric stuff pretty well. We have a business which is heavily metrics focused and monetizes very well. Why don’t we do business together?

I’m like…I said I don’t think there’s enough money in the world for me to work with you. And they might have had a business that involves watering virtual cabbage. [laughter] And the guy misunderstanding my objection said we can be arbitrarily generous. [laughter] I’m like no. I think there’s like literally a biblical line to the effect of: What does it profit a man to inherit the world and lose his soul? [laughter] You’re like hitting, asymptotically approaching that for me. And they’re like, so you’re saying you want stock options? [laughter] Bit of a clash of cultures there. That one didn’t work out. Worked out great for me. So expect to pay through the nose if you do this.

Also it’s virtually impossible to hire these people. Not bragging but just telling you a state of the market, one of my consulting clients they saw we had made a couple of million of dollars for them in the course of a two week engagement. And he just like tried to annualize my consulting rate and said alright I want to hire you to do this full-time. I’ll offer you a salary of $700,000 a year. And I said that’s very flattering but I really love running my own business, so no. And he said $700,000 is probably more than any engineer who’s ever been offered to do this.

I’m like I’m not bragging or anything but that’s actually sort of on the low end of the spectrum at the moment. It’s like oh my God! It’s kind of like sales guys for enterprise software, like top performing guys at Oracle. The don’t do poorly for themselves. And this skill set gives you…like you can verify with SQL and by talking to the accountant exactly how much money this person made for your company last Tuesday. So yeah, it tends to scale up on the competition. So I would suggest rather than hiring proven talent, build your own talent or take maybe a chance on somebody who’s a little less established in their careers.

Side note. Career advice. So, everybody who runs a business just go la-la-la for the moment. But for the young engineers in the room. If you just spend a few weeks working on just say the MailChimp API or learning how to run A/B tests. Maybe spend one week writing a A/B testing framework in your tool of choice. The net present value of your future career takes a double digit percentage jump just because rather than going into your next company and saying, or your next salary negotiation and saying yeah I was in charge of optimizing SQL queries on the customization screen for one of our products that very few customers ever see. You’ll be able to put specific measurable accomplishments next to your resume like your product manager is doing. Like I was the individual performer in charge of increasing our conversion rate from 12% to 16% which increased sales of the company by 1/3rd. And if they say what are your salary requirements? You can say something like, you throw out a number and I’ll throw out a bigger one and we’ll compromise on my number. [laughter]

Anyhow. So I’ve been talking about the strategic aspect of this cause I know Business of Software is a very strategic oriented conference a lot of folks here are business owners rather folks that are in the day-to-day. But I know many of us including myself are more tactically focused. So I wanna talk about just one tactic under this broad developing marketing thing that you can try in your businesses and over the next couple weeks. It’s kind of one of those secrets that they’ve known for forever in one end of the software industry. We haven’t really heard of it yet in the small SaaS market. At least I started hearing about it a lot about a year, year and a half ago and I’ve seen a lot of people do a lot of interesting things with it. It’s called Concierge Onboarding.

So there’s a fundamental mismatch between r-value proposition as we understand and r-value proposition as the customers understand it. We think we’re selling a software product and the customer has absolutely no care. They do not want a software product. They have this problem, this burning hole in their live and they want a solution to that problem. And they only want the software product in as much as it is a partial solution to the problem. But it’s almost always the partial solution to the problem. But it’s almost always a partial solution. So we want to get things where it is closer to a full solution. At not like a solution like, Enterprise Solution Provider. Because that just mean you’re selling a product but tacking a lot more zeros onto the invoice.

Here’s a traditional SaaS payment grid. I have picked from my software company but you could pick from any of 2,000 at this point. We tend to segment the market and segment prices based on, you get a little product for a little money or if you want to buy more product you pay more money. And that’s made a lot of people a lot of money. It’s sort of an unsophisticated way of segmenting the pricing where the customer is coming from. If the customer is more willing to solutions, more willing to pay to things that get them closer to the outcomes. Maybe we can think of segmenting software pricing a little more like this.

So we align the pricing with customer goals. So if we get something that gets you a little bit of the way to the goal like hypothetically we’re just gonna tell you what you need to do. We’re not going to give you a tool to do it. We’ll charge you a little. If we give you the tool to do it, we’ll charge you quite a bit more. You’re almost there all you have to do is use it. But if we actually…we’ll give you the tool, then we’ll actually run the tool for you. You just write us the check we’ll take care of this. Then you can get away with charging gastronomical amounts of money. Gastronomical is not the right word. I’ve lived in Japan for ten years. [laughter] I don’t speak the English anymore. Anyhow, so how does this relate to the SaaS pricing grid we all know and love? You couldn’t put things that…say you have a customer that would numerically fall into one of the lower valued plans. You can offer them, plus a bit of help getting started, bit of help with this Concierge Onboard that we’re gonna talk with. And only if you’re on one of the higher tiers and then use that push towards the ultimate success to justify them moving up to one of the higher tiers prior to starting using your software. Whereas just the numbers are just the feature set. Wouldn’t have convinced them to go to the higher tier.

Let me tell you what Concierge Onboarding is in a little more detail.

So you reach out to the customer either when they start the trial or when you have a sale or a little bit before that. You say look we know it’s kind of hard to get started running with the new software product, so we’re gonna take care of all that stupid nonsense for you. Like importing your stuff, configuring software, doing the training. We’ll take care of that. We’ll handhold you through this so it won’t feel like oh my God this is one more thing of work I need to do today. It’ll feel like we are in your corner on it. If you just do this in a totally automated fashion people are pretty good at blowing off a machine that says you haven’t set up users at your organization yet. Would you like to set up users? And the person will think I’ve got three things to do today maybe if I get done with my entire to-do list I’ll set up users but blah.

But if it’s a helpful person who has been talking to them for the last couple week on saying hey Bob, just circling back to you. I remember we talked last Thursday and you said you were gonna set up the users. I appreciate you might have been busy. Is today a good day? Or should I call you back this Thursday? They’ll think oh shoot, I did promise Nancy that. I don’t want to let down Nancy. She’s been so nice to me, so I will set up the users and I will get successful use in this software. Which I know I wanna do. I asked you for the trial of it after all. And then it becomes much more likely that the sale happens.

So here’s a hard requirement for Appointment Reminder and I’m sorry it’s in really little letters. I sold to customers across a wide variety of businesses who already used things that had customer data in them. So customer name, customer phone number. They don’t want to hand-type 2,000 customer’s names and phone numbers into my solution. So they will say I’m a dentist, do you interoperate with Dentrix? I’m a dentist, do you interoperate with Othrodontrix? I’m a dentist, do you interoperate with this? I’m a furniture mover, do you interoperate with this? And I’ve literally been asked this about several hundred systems.

So I had this bright idea as the product manager. I’m going to create this awesome totally automated system where a totally non technical user will able to import a file from an arbitrary format. It will give them a Google Docs flavor user experience where everything is nicely laid out into columns. They will select which columns map up to my database schema -while telling them what a database schema is- then they will hit go and no matter how terrible that data is it will work with no problems. And then I will not have to write conversion scripts for all of these things. And then I started actually building this and I’m looking in jQuery plugins to get that wonderful Google Docs experience in just two lines of code.

And after about two weeks of banging my head into the wall like oh God it would be quicker if I just typed in all the data myself. And I thought hey wait a minute. That could almost work. So I put a little button in software that said import your contacts and if you click it, it says email us a file in any format and we can import it for you. Can anyone guess what the implementation of this is? Pay no attention to the man behind the curtain. [laughter]

So I say it queues for 24-48 hours. It depends on whether I’m at Business of Software or not. I’ve done three of them in the last 48 hours or so. So I just go in and write a little Ruby script to grab whatever the Excel file or comma separated values and munch it into things that my Rails program can understand. And I’ve done file formats I didn’t even know existed. Excel, all the versions of it that Microsoft sells and some they probably don’t. I also had…so one company probably shouldn’t name them. Hyper compliant solution with enterprise grade encryption that even the NSA couldn’t crack. Now I’m not sure about the NSA’s internal capabilities, all I’ve got to say is that all your base64r [00:42:47] belong to me.

But did get that data imported without having the customer have to retype it. So this…did this work out on conversion rates for my trails? If you take me up on the offer of I will import all your data for you, you are four times more likely to actually pay me for at least one or two months of using this software. That was pretty cool because this generally takes me five minutes of work if it’s just a simple Excel file or maybe an hour if I have to write a full featured file parser for their oh my God complicated Excel, sorry, XML madness. Works out pretty well.

Also turns out that guesstimating based on my turn rate these customers have about a four times higher lifetime value. Because it typically tends to be the larger customers who are on the higher level plans. They turn less and they’ve been paying over more money every month. The nice part about that [00:43:37] is that they are multiplicatively effective. So let me just throw out a random representative number. If the value of a free trial of Appointment Reminder is about $100 probabilistically. The value of somebody who clicks this button in my software is in the thousands. So it’s totally worth the 5, 10, 20 minutes it takes me to do this for them.

So that’s just one example of a Concierge Onboarding thing you can do a lot of them. I’ve seen wonderful things from different companies. Drip which is an email marketing company, they said the big problem with an email marketing tool if you’ve never done email marketing before is oh God, what do I write to my customers?

And they say look do you have a blog already? Just point us at your blog, we’ll have an actual human being take a look at your blog and distill some of that into an email course for you and then you just have to click go. Or we can click the go for you.

WhooshTraffic does marketing analytics. A lot of companies don’t really understand where marketing analytics plays in like, how does that actually make me money? Their CEO was like me a consult on this for a number of years. She said look sign up for our service, I will hop on call for you for two hours and write you on a marketing strategy. Totally free. As long you sign up for the plan that is at least $1,000 a month. And so she has a lot of customers paying her $1,000 a month of totally free. And of course it doesn’t scale for a venture funded business if you’re doing two hour consults but because the vast majority of the billing comes at 95% plus marginal profit on the license or the SaaS billing rather than on the low margin services business. It works out much better for her investors and for her.

PlanScope project management. Thought I saw them over there. Hi Brennan. Brennan said if you are getting started on this I will walk you through it on Skype no questions asked. Well lots of questions asked like what do you like about this? What do you not like about this? It’s a great opportunity for both figuring out what is not working about your software and also for getting individual customers moved towards successful use of the software in their businesses.

Campaign Monitor email marketing. They will assign you a dedicated rep and that person’s job is, you know, they are handling 20 people in parallel but I think they are called a success coach or something. And they say hey Don’s Bakery lets talk about your email marketing needs. He Don did you write that newsletter you said you were gonna write about awesome things happening in the bakery for Halloween? You need some pointers on that? I’ve got some great ideas on how it won’t spam filters and we’ll totally sell more of you delicious cupcakes. And they handhold until the point you can successfully send that out to your email list and then they get back in touch with you. It’s like awesome job with that email, 200 people opened it and 30 of them visited your website. Tell me if you sold more cupcakes today? And the customer has an awesome experience out of that.

And I really like how this feeds into sort of like Innovation Cycle kind of like the lead start guys are always talking about. So I’ve often talked about doing Onboarding for customers in code, in product. So building out like a custom tour. It takes a lot time to do and tours are sort of brittle. It’s kind of hard to change on the fly. It’s limited by engineering. But Concierge Onboarding is very easy to change on the fly. All you have to do is change your mind. So if you want to emphasize a different feature of the product, talk about that. If you want to emphasize to try a new way of explaining things, try that. Learn from doing that quickly. Then you can build that into the Automated Onboarding.

You can take the stuff you learned from the Automated Onboarding which has nice rigorous methods and is A/B testable up the wazoo. Push that into the product proper and then have the product inform the design if you’re on the Concierge Onboarding. So this tactic pretty awesome in general.

So a great way to get started with it, much like an earlier presentation. Offer it in a limited or timeboxed fashion just to figure out how much this is going to cost you. So you don’t have to put a publicly available thing out on the website saying we will import all the records from all the users. Just reach out to five users who signed up today and say hey, I’ve got a little extra time I want to make your life easier. Send me a file I’ll do all the importing for you. And after you’ve done this a few times, you can control how much work that commits you to. But you can figure out okay, is this going to take me one hour of time? Is this going to take me ten hours of time? Yadda-yadda.

Figure out where it makes sense economically and then maybe roll that out to just the higher value plans to use it as an incentive for people to select up into the higher value plans. And then if the economics work at scale there’s no reason you can’t do it across all the plans. I know a lot of companies that do. Like Campaign Monitor they’re average customer is paying something on the order of $30 a month and they have an actual human call you on an actual phone to talk about your actual email every single time because they know the metrics are there for them.

I have a different topic that I’d like to close this talk about and it’s not about marketing at all. A shout out to Greg on this one cause I kinda feel I need to say it publicly. There’s a great phrase in Japanese that says I would like to take this opportunity to borrow this space and talk about something. And I’d like to borrow this space partly to be the introvert talking through things out loud and partly because I think it might resonate with some of you. And maybe for the folks it doesn’t resonate for you might see something in folks you know. Maybe that will be useful for you, I hope.

So when I had proposed to Yuriko I said look there’s something I need you to know before you decide on this getting married thing. About every six months or so I have a blue period and I get a little down for about two weeks. My personality changes a little bit. I’m generally not a very fun person to be around and then I get better. And then I’m my nice normal self until the next time it happens. I know a little bit about this but stress tends to bring it on. And not so much. But thought you should know because I think this is important. Also I never tell anybody about it. Prior to this presentation I think my wife, my brother and God were the only people who knew. She said I think there’s a word for that. I’m like, the D word? I don’t like the D word. I like another D word, it’s called denial. [laughter]

And so something happened to me recently. This August, you know the random ups and downs of running a business. I quit consulting this year. That caused a large cash flow shock to the business. The SaaS didn’t grow quite as fast as I was expecting it to. I had what is in the grand scheme of things a very minor business problem of the type that we all have on a every couple of months basis. And normally I would just go on through this but it made me feel a little on the bluish side and I made a very aggressive schedule for getting out a new product to reduce the cash flow deficit and get back to goodness. And overwork pushed me a little bit deeper down the spiral and then there were problems with the product that pushed me deeper down the spiral. I became totally unable to do creative work on the writing or marketing aspect and that pushed deeper down the spiral and then six weeks just vanished.

I think I did like one day of productive work in that. There’s a funny thing about the D word. Depression. When you’re in it you don’t necessarily know that you’re in it. I like to think I’m a fairly smart guy and I have a bit of introspection. I had no clue this was happening to me in August or September. No clue. Not the foggiest. My wife was telling me you look a little run down. I’m like i’m a little stressed by you know, the work situation but you know, it’s a stressful environment and it’ll get better after I remove the stressor. And she was saying things like when you get stressed do you normally like stay up ’till 5AM playing League of Legends and then sleep until 4PM? I’m like, I don’t normally do that when I’m stressed. I didn’t do that now. She’s like honey, it’s 4:30PM in the afternoon and we’re eating breakfast together. [laughter] I’m like, really? And there were a whole lot of like, oh really? It’s amazing it’s like Dr. Jekyll and Mr. Hyde. But Mr. Hyde thinks he’s Dr. Jekyll because…Maybe to connect to the engineers in the room I’ll use a metaphor. If you were an operating system you execute on untrusted hardware. [laughter]

And the float control of what you think you are is dependent on your brain not lying to you. And unfortunately your brain can lie to you. And I was thinking symptoms are norm- all systems are go. Business minor trouble. No problems. I think, what was it…My wife came up to me one day and asked that if we had…deeply personal stuff coming but just say no…she asked if I was dissatisfied in our marriage. I’m like oh God no! This is, marrying you was the best decision ever! I’m so ludicrously in love! Why would you think that? And she said well you’ve been really withdrawn for the last few weeks. You barely talk to me. When I talk to you, you don’t listen. Rather than getting up in the morning with me and going to bed at the same time as me, you do basically none of those things. And the last time I asked you out on a date you said I’m very tired I wanna stay inside and watch The Wire. Which by the way, The Wire, excellent television. Not a great depression anecdote…antidote. [laughter]

Anyhow, and I’m like oh, oh, you have even less introspection on my brain than I do. I need to communicate that stuff. And so when I was talking about this after getting out of the blue period when I can look back in hindsight and say oh yeah…that was the little D demon sneaking up on me. One of my friends said I would totally taken the train out to, from Tokyo over to your town. Three hours. To help you with that if you had said anything. I didn’t think of saying anything because I didn’t notice I was in it. I didn’t want to burden anyone with my problems. And he said well you’ve got your best friend Keith who lives two blocks away. I’m like I kinda assumed Keith would be busy every day and didn’t talk to him or call him out. And besides if I called him out that would have required a whole lot of energy. Like getting dressed and getting outside and getting back inside and getting undressed. And that’s just something I wasn’t ready to deal with.

And he’s like if you ever feel like, this is a warning sign. So talk to somebody you trust. You can talk to a professional. Unfortunately in Japan they were certifiably Medieval on this so professionals not so great. But your friends are there. If any of you in this room ever need somebody to talk to who’s been through it, my contact information is readily available. I will drop what I’m doing to talk to you about it. Okay and contact information, here’s my blog, here’s my email address, my newsletter. If you enjoy some software you should probably be on it.

Thanks very much for your time. [applause]