Evolution of a SaaS Marketing Team + AMA | Andrus Purde, Outfunnel | BoS Europe 2017

Andrus Purde, Founder, Outfunnel

People talk a lot about scaling SaaS, or scaling your hiring or culture. But what about marketing? How does your Marketing need to develop in line with the rest of the business? Andrus talks about the four distinct marketing stages your company will go through as it grows; and how, ironically, just as you get comfortable doing something, you’ll need an entirely different skill set to continue successfully.

As an Estonian, he thinks a lot about conflict, (his words!). In this talk, he likens the four distinct stages of SaaS marketing to the evolution of war.

  • Hand-to-hand combat
  • Scalable weapons
  • James Bond villain stage
  • Weapons of mass destruction

also available on the podcast

Video, AMA, Slides, & Transcript below





Learn how great SaaS & software companies are run

We produce exceptional conferences & content that will help you build better products & companies.

Join our friendly list for event updates, ideas & inspiration.

Unsubscribe any time. We will never sell your email address. It is yours.

Slides from Andrus’ talk here


Andrus Purde: Thank you, Mark! I think there was a slightly apologetic undertone with the length of the talk. For the Americans, I live in Estonia. It’s between Russia and Latvia and Switzerland. If you live near a country that’s run by Putin, you’re always prepared. So when I got this opportunity to speak here, I had a 3 minute version, 7 minute, 20 minute and 3 hour version. This is the 3 hour you’re seeing tonight.

Bit on myself before I begin. I’ve been in marketing for just over 20 years. The last 11 years in technology and start-up marketing including 4 years in London. I lived literally around the corner from here. I was gutted to see that my favourite kebab shop has been closed down. It’s now called handmade gourmet kebabs. I’ve never seen these words in the same sentence before. And for the last 24 hours I’ve been thinking is it good marketing or bad marketing? But yeah, some good and bad marketing coming up for you in this talk.

Pipedrive is a sales system. Which makes it very easy to get organised in sales and close more sales opportunities. Some important numbers about us, we’ve been around for almost 7 years, founded in 2010, we have more than 50,000 paying customers, we have raised just over $30 million, there’s about 280 of us in the team and this last number as many people guessed the registration number of the UK which was founded last month with a view to setup the marketing revenue creative office here.

Evolution of Pipedrive Marketing

And I’ve had the same job title at Pipedrive, head of marketing, for the last 6 years. But really I’ve had 4 very different roles and contexts with 4 different sets of challenges. Before I tell you about this journey, I wanted to apologise, sincerely, for using a war metaphor. I don’t really think the customer is our enemy. In fact I think the opposite is true, but I haven’t been able to find a better metaphor for this so please bear with me. And if I can think of a better metaphor, maybe Mark will invite me back in a year or two. Same talk but with a better metaphor.

Hand to hand combat

So the first stage of Pipedrive marketing wasn’t like marketing at all. It was the founders of the company asking, begging, threatening their friends and business contacts to add their data in this beta product. And we had 400 people kind of agreeing to that, 100 people actually signing up, about 40 people putting the data into the system and as we exited, we had 18 paying customers. We were in business.

Scalable Weapons

The challenge was at this pace it would have taken us 90 years to do enough to start paying salaries so we needed to do something to get more customers. We did a promotion, we got our first kind of PR cherry taken. And then each of these things got us a handful of customers at the same time which is great if you’re a government start-up. But all these things were one-offs. I really need to find scalable ways of getting customers. In the lingo of this talk, we needed machine guns.

The Pripedrive Marketing Templ(at)e

Andrus Purde SaaS Marketing

And as we were starting to look for scalable channels, for a long time I thought of Pipedrive’s marketing or just marketing as a Greek temple. When I say that, I don’t mean that the marketing was in ruins. Or that marketing was managed with the efficiency and diligence of the Greek finance minister. What I meant was, like a Greek temple, my marketing has a solid foundation, it has pillars and they hold something valuable up. In the case of Pipedrive the base, the foundation was a good understanding of the customer, of the value proposition and the brand. And the pillars, we tried many more pillars to come up with these four where SEO and content, it was paid, it was what we now called growth engineering, but it was basically a team of engineers who were tasked with the integration and connection to the software and attract to our program. Of course, these things held up our customer base and we did a bit of customer lifecycle marketing and engagement marketing.

Scalable Channels

Getting to these four pillars like I mentioned was a lot of trial and error, mostly error. In hindsight if I could turn back time, I would frame our first couple of years marketing differently and I prioritise our experiments to be more effective. Who here has read the book by Jim Collins? So you know the hedgehog analogy – doing one thing really great. As a really early stage start-up it’s not enough to do one thing well, you have to do two things really well. I invented the two hedgehog growth model. And the two of them are – whenever I talk to customers of Pipedrive or of other apps, I ask where did you hear about us or that brand? And two answers come back as dominant. One is I just heard about it from a friend or colleague and the other answer is I just googled and searched online. So I think the key to testing for an early stage company is to work on those recommendations and focus on the visibility. I like the word findability, which I think, is more precise.


And with recommendations, it just has to, you need to have a great product. There’s very little marketers can do about it other than build a program and add incentives to try and encourage more referrals. Just to double-check something to having a referral program I emailed our customers who had used the program to ask why did you do it? Why did you recommend Pipedrive to a friend or colleague. Overwhelmingly, they do it because they just want to help somebody or they do it because they like the service. We offered 3 months as a reward and these were almost not on the radar at all. Over the years I tried different incentives and optimising the referral program. The trick is to not over do it, to keep it a recommendation a referral not a transaction. If you give too much it, we found the effectiveness of the program goes down, not up.

And another quick tip, it really makes sense to be early on. This is successful invites sent for the referral program. There is no better time to ask for referrals than the first 1-2 days when people are still fresh and they haven’t found any flaws with the system yet. I think same is true for marriage and many other types of commitments as a married man.


And then the second part of the two hedgehog growth model is the findability. Which is not really a channel, it’s the web of channels. On the front end it seems like a bunch of tactics, like there’s SEM, there’s SEO, concentration, review management, PR, maybe community. And on the back end the process is fairly straightforward. It’s understanding how the users make the searches. Usually it’s google but it can be YouTube or an app store. Your friendly neighbourhood search engine. Then it’s understanding what these people are searching for, what the search queries are, what the popularity of each query is and the comparativeness. And then there is analysing what comes up in these places where people search. So looking at the first 10 search results for google for instance. And then being omni present, so no matter what customers are searching for or where they click, you would want people to find you. One example, in a CRM world, like I am, it would make sense to be lifted for a CRM software. But it’s very competitive and it’s a very expensive key word. It’s not reasonable or even possible in these terms for us to be in the first top 3 of google analytics. But if you search for CRM software you will probably find us no matter where you click on the first page of google.

Time to Mobilise

And then as well, as we were building out these channels, again this seems like it’s a lot cleaner and simplier in hindsight. It was a lot of trial and error of the process. But we were lucky enough and smart enough to be able to find a couple of channels that worked for us and we quickly found out we were running out of hands to man these machine guns. Or put another way, the machine guns got so complex, and unique, specific skills to man them. So we started a general mobilisation.

For a while, every problem I had looked like a hiring problem. Every opportunity we had looked like a hiring opportunity. I always thought of myself as a marketer but I would put my hat down and focus most of the time on hiring and on boarding. By the way we’re not burning books here, or voodoo dolls of competitors. This is a team event at the first hiring, we did fire walking as a team building exercise which was interesting. Nobody got burned. Although I made, and we made, a lot of hiring mistakes, we were lucky enough we got the team growing rapidly. But instead of going faster, we started to go slower. People started to bump and there were things done twice and things not being done. I think once again, I can put off my hiring hat and put on a process design hat. Peldi mentioned yesterday that this was in retrospective. I spent the last year probably tinkering our monthly marketing retrospective process and the work is still ongoing. The last stage is about designing the systems, processes and communication. We have to make sure the people don’t bump into each other and have their swimming lanes cut off for them.

The Bond Villain Stage

Andrus Purde SaaS Marketing

So this James Bond villain metaphor now. When the villain presses a button some amount of troops go and conquer a country or a planet. They have their people and processes figured out. And similarly, I think a well functioning marketing team can press a button if they want to press one when they want to launch a new product. Unless of course James Bond runs in and ruins everything.

Weapons of Mass Distribution

So we’ve gone full circle in my journey of hand to hand combat where we needed the customer development skills and making sure we have product market – it’s much more like sales than marketing. Then there’s a brief stint of marketing, this scalable weapons stage. Then in the 3rd stage we were really finding the people and we needed to be very good at finding and onboarding and sometimes firing people. And in the last stage which never ends is tinkering with processes and systems and trying not to kill the culture in the process. Which again, looks much easier in hindsight.

Fog of War

The thing is it’s difficult to tell when one stage ends and another one begins. No one shouts hey guys, this stage ended now, let’s prepare for the next one. In military terms, there’s even a term for it – fog of war. When you’re in the battle and you don’t know where the enemies or your own troops are. So in the fog of war it’s very easy to be in the wrong stage and it’s costly. The mistake I made and I’ve seen in other start-ups is trying to be in a stage too early. So it’s not a very smart idea to start building a team before you even know what channels you need to manage. Or it’s also tricky to start looking for scalable channels if you don’t really have a product or a product market fit. You can’t win the hand to hand combat with a hand grenade. Nobody walks away from that fight.

And the mistakes I’ve made a couple times is being stuck for too long. I took too long to start to focus on hiring, I tried to man too many machine guns with my own two small hands. And I started this process of systems designs too late. But having gone through this journey once I think it helped me identify the signs of when does one stage start and end better and I hope some of you will be able to tell this or see if the signs after this talk. Any questions and comments?

Learn how great SaaS & software companies are run

We produce exceptional conferences & content that will help you build better products & companies.

Join our friendly list for event updates, ideas & inspiration.

Unsubscribe any time. We will never sell your email address. It is yours.


Mark Littlewood: Ok, could you go back to your four pillars slide?

Andrus Purde: Yeah.

Mark Littlewood: That one! What am I missing if you’ve got 20 and 10 and 5 and 8 as your top ends of growth, where is the other growth coming from?

Andrus Purde: I had a marketing thesis, the only marketing product which don’t need marketing. The rest is coming from organic growth, people recommending us. This last part is referrals which are checkable and trackable in the program. The big one has we have most the skills coming from the product and strong word of mouth.

Mark Littlewood: So it doesn’t matter, you get 50% of the growth without being there.

Andrus Purde: My job was very easy.

Mark Littlewood: Ok, that’s very interesting. Other questions?

Audience Question: You mentioned about referrals, that you experimented with why people did the referrals for goodwill or having a trigger, did you try not offering anything at all and seeing if that had a positive or negative effect?

Andrus Purde: We haven’t tried it, no. I think it’s incentives, but it might not be a bad idea to remove incentives altogether. We have a referral scheme where both the recommender and recomendee receive some free Pipedrive time. Talking to other marketers, that seems to be the best practice. We tested not having referrals at all.

Audience Question: You talked about growing your team. What have you found the best way to find good marketers? Cause it seems to me there’s many bad out there.

Andrus Purde: I’m still searching. I think the mistake I made at least once which cost us a year of team ramp up time is finding a good marketer and not talking to the underlying expectations. What business KPIs do you expect to influence? So finding good people isn’t easy but finding the ones who have the same underlying belief is really important. Now I will sit down and have a painfully long discussion about the very deep expectations and for a function before I send out an offer.

Audience Question: Out of the 4 pillars, 2 are fairly easy to measure but the growth engineering is pretty hard to measure exactly, how it affected growth. How do you do that?

Andrus Purde: You can measure content very easily but google doesn’t always tell you what the search queries are which people use to land on your site. When you use google, a web console to get the query and there is some calibration you can do to measure that very precisely. And this is everything we do which is trackable to signups so each of the teams we keep in that pocket and it drives the traffic and if it drives new traffic for us we can use tracking codes. I do agree that sometimes doing new features and seeing it it’s difficult to assess especially if we don’t have millions of users but everything we do in that team is measurable.

Audience Question: Hello! In our growth and marketing, we did 1-2-4 and we’ve skipped 3 so far in terms of hiring people. So we’re super far behind in terms of that. We’re under-resourced which is similar to what you guys did before. How did you prioritise which hires you needed to do first?

Andrus Purde: We started – I was just a marketer in the beginning. So we started hiring based on the things which basically worked for us. We were experimental and opportunistic. But there seems to be more potential there which you didn’t have funds to hire more people, we just hired where we thought it will have the most returns based on the earlier results.

Audience Question: If you find you skipped a step, what do you recommend?

Andrus Purde: To go back!

Audience Question: In your 1-2-3-4 stages not this slide but your overall deck, if you find that you have missed stage three and you’ve just gone 1-2-4, what would you recommend?

Andrus Purde: I’m now speculating because I haven’t skipped a step myself. I would think it just maybe it’s wise to do 4 before you do 3, maybe smart people will do that because then by the nature of 3, if you choose successful with 3, you have to do number 4. If you do 4 well, maybe you don’t need to hire as many people. So I think you can go both places. I think it depends on how fast your growth is and how much resources and time you have.

Audience Question: Regarding your, the 2nd column of the Greek temple that was paid advertising. What channels have you found worked well? Have you tried AdWords and Twitter and other things? Which one would you recommend based on your experience?

Andrus Purde: I would recommend whatever comes up if you think of the hedgehog, findability. So whatever keywords are coming up for you, look what sites comes up if you search for your keywords. Some sites are paid sites and you should focus from these paid sites. We use Facebook now and it drives some results. So like among them I will search to AdWords and some directories. They have different categories.

Audience Question: Sorry, one other question on your pillars. 80% of your budget goes on pay. In our experience we found we have much higher conversion rates from SEO than organic paid. So we tried if we can migrate our paid traffic unto SEO. Doesn’t go away and it seems the more effective – have you had any experience with that?

Andrus Purde: Of course! We see the same keywords that worked for SEO worked for paid and vice-versa. So if we had a lot of time and no funding, we would try to do the same. Since we are reasonably well funded we can afford to be present in both places, and just take more chances and then no matter where the customer clicks, they’re more likely to find us.

Audience Question: I feel like I had a Pipedrive advert on a podcast. I wondered how you got on with that.

Andrus Purde: Podcasts is actually not a dominant channel for us, and this 2 hedgehog model works at the early stages. When you start building the brand, things become more tricky so they are more difficult. Some of them are purely for branding. We can measure them when people select the choice when we ask where did you hear about us? But this is an indirect way of measuring podcasts. We also do some which are trackable. Go to Pipedrive and we get an offer. We try to triangulate these two sources and we use podcasts for both short term region and also for the branding.

Mark Littlewood: Right on time! Thank you very much indeed!

Learn how great SaaS & software companies are run

We produce exceptional conferences & content that will help you build better products & companies.

Join our friendly list for event updates, ideas & inspiration.

Unsubscribe any time. We will never sell your email address. It is yours.