Tim Barker has established three startups and has seen the ups and downs of the startup journey, from his first business being acquired by Salesforce – where he spent five years in the formative age of cloud computing – to his current company DataSift being acquired by Meltwater in 2018. He is a keen advocate for transparency, eating humble pie, and being big enough to own your mistakes.
In this cathartic talk from Business of Software Europe 2018, Tim talks about some of the lessons learned he learned during some very, very tricky days for DataSift. How do you lead well when disaster strikes your business? How do you motivate and lead a team when the future is uncertain? Tim shares his experience of leading DataSift through an unexpected nightmare, and the importance of uniting your company around shared values.
Learn how great companies are run
At BoS we run events and publish content that is highly valued by anyone trying to build, run, and scale a great software company.
Sign up for a regular dose of actionable and useful content:
Tim Barker: Hi everyone. This is the first time and probably the last time I’ll ever give this talk. Hopefully It’ll be informative for you. It will be therapeutic to me. I will talk about managing through good times and bad times so don’t worry there will be good times before I get to the bad stuff. Being British we like a bit of bad news. Part of this session is quite sensitive; I’d like to speak quite freely so we are going to have one part of the session with a redacted slide. We’ll turn the cameras off and we’ll chat just attendees only. For those of you watching the recording stream in the future, you’re about to miss a phenomenal event. And when it gets to a safe part in the deck we will turn the cameras back on and it’ll go from me being a professional you know rocket fuelled entrepreneur to like a nervous wreck later on in the deck. They’ll be wondering what happened.
As you know how these things work when you’re putting a presentation together you think what’s the one slide that can summarize the best of times and worst of times. So this is me when I went to Disney with Greg who was 5 years old (he’s 16 now) and we just walked through the golden gates of Disneyland and of course as you know if you’ve been there as you walk down main street there’s a whole load of candy stores, temptations. And so Greg was immediately drawn to “let’s go and get some candy, Dad!” and I went “No no we’re going to do that on the way out”. And at that moment that’s when the official photographer took a picture of us. You can see here his cross armed angry face. We’ll revisit this later because there is a rule in life – I’ve got another story to tell about that.
But a bit about who I am. I’m Tim Barker the CEO of Datasift. It was acquired in January by Meltwater so we’re now part of a much larger organization. I’ll touch on that but I’m saving that one for another talk track which I hope to come back in a couple of years and talk more about. So I founded 3 startups – two of them have been acquired as I say Meltwater acquired DataSift and a decade ago one of my startups called Koral was acquired by Salesforce so I spent five years at Salesforce at the formative years of cloud computing when it wasn’t a dead certainty that it was going to be the future of software. But my personal journey is a developer to CEO. I’ve got some credibility on this but I’m not going to talk about AI or Blockchain today cos you’ve probably had enough of that. So as well as sort of thinking about themes for today’s session you know this is going to be about lessons learned and I thought “Who’s the best person that’s ever gone through hardship?”. And I immediately thought of Ernest Shackleton but then I saw this advert and I thought maybe I won’t feature this because nobody sets out on a mission for hardship unless you’re Ernest Shackleton. For those that can’t read it on the back it says “Men wanted for hazardous journey, small wages, bitter cold, long months of complete darkness, constant danger, safe return doubtful, honor and recognition in case of success”. That’s the entrepreneur’s mission summarized there.
Of course we all start with bigger ambitions – to nail it, to crush it, to knock it out of the park and all those other colloquialisms we’ve come to love. But let me tell a bit about DataSift and where we started. For those of you that don’t know, DataSift launched efficiently in end of 2011 with a proposition that was pretty hearty. We were partnered with the world’s social networks. All of their data would run through our technology because at the other end of that there are brands and businesses that really wanted to understand what was happening on social media; what was happening around their brand, their industry, their market. And the reason that we started on this proposition was as one of the speakers mentioned earlier – there’s no straight line to success. We didn’t start with this proposition. We actually started with a proposition to build a news site that was powered by Twitter. And so when that plane landed on the Hudson River back in 2009, the news broke on our website called TweetMeme. And in order to get more news onto that site we invented the Retweet button. when you saw a news story you want to share on Twitter, you press the button publish onto Twitter, go through our service. What we found was that we thought we’d have companies flocking to us to advertise on our algorithm driven news site. And we did have companies come to us. They were like Dow Jones, Bloomberg, BBC and they just wanted one thing: They wanted the data that went inside these kind of algorithms.
So we pivoted to what became DataSift and I got to say it was a wonderful ride for the first few years because I think it had three key things three lessons that I learned from five years at Salesforce. There’s three things that you’ve got to do to escape velocity essentially on a fast growing business. The first one is market relevance. We had a really high market relevant problem, you know DataSift launched just after Arab Spring. Remember when social media was going to be all things about positive change for the world and people had big banners that said “thank you Facebook”? Doesn’t quite work that way these days things have moved on a little bit. But there was a huge amount of interest and relevance around this. Yes, front page of the BBC news site when we launched our partnership with Twitter and we became the world’s first company to allow you to do historical analysis of what happened on Twitter. Yes, we did confuse some people – one of my biggest moments as a marketer was having Joan Collins angrily tweet towards me about some privacy breach. She thought that her tweets were just for her followers. But me and Joan are friends now so we’ve healed that one. And then we obviously then use that kind of anchor tenant of a partnership to then grow additional partners. We partnered with Tumblr, WordPress, bit.ly, a whole load of companies that had public data that we can help other companies gain insight from. We had high relevance. We also drove a huge amount of inbound. This is a B2B product. you know with a small team doing about a thousand leads a month of some quality was an impactful business model. And we made it low friction because our customers were developers, ones that might want to get their hands on the data, start working with that building analytical products. So we gave we had a seed and grow model: so they could start with a very low commitment of $10 and they can scale from that. And we still have customers today that started on $10 plan and now spend perhaps $50,000 a month with us.
So it was really low friction to bring people into the funnel, convert them, and scale the revenues from there. And over the first three and a half years we did exceptionally well. You know when you look at the trajectory of high growth companies especially in B2B enterprise, I think we rank up there pretty much at the top of them, the top right. We went from zero to $25 million ARR in like three years and a quarter. We raised a ton of money as part of that. We raised $52 million along the journey, often just by really showing the strong growth metrics from that. We had and we grew a massive team. We had engineering in Reading – anyone from Reading in here? Thank you, sir, it wasn’t even a shout was it, we’re modest in Reading. we raised a good amount of money. We had teams in Reading, San Francisco, a team in New York. A team in Canada as well, an Inside Sales team. We had a huge amount of growth behind us. And I think really high expectations of three four years forward we’d be like up to $ 100 million. And that’s the magic number for IPO.
And now for those of you watching on the recording unfortunately we’re going to park this conversation here and we’re going to pick it up in a few moments once I’ve unloaded another chapter of the story to all of you in the crowd. So cameras off. Okay…
I know. I know. Crushed it didn’t we, crushed it. That was insane! Those are the bad times and now I’m just going to talk about good times again. there were perhaps three or four other things I want to share with you things that I learned as we dug our way out into this awesome opportunity that we had. And again some of these things are things I’ve learned in the past but perhaps haven’t really applied as strongly as we needed to.
So hands up in here. Anyone apart from the Reading guy heard of this thing called the V2MOM before? One person, thank you sir. I spent five years at Salesforce and the aim of the V2MOM is to get the whole company and every employee aligned to the single plan and that plan fits on one page. This is something that was really, you know, was drummed into me over five years which means that the CEO has a plan and the plan is a plan of intent not a plan of execution; meaning that we know we want to get to but it’s not about micromanaging it to say here’s how we get there. It sets the end goal of what we’re trying to achieve. The five things in the V2MOM: Your vision which is really the call to action. Where do you want to be a year from now? The values, what things are critical to get there right. Those values might be you know might be customer success might be your key value. It might be around kind of delivering on the promise of what we might be building. They’ll be different for every year and they really are things that you need to make sure are front and center for every employee. There’s what’s called methods which is the stuff – most V2MOMs start by writing ‘the stuff’, the methods first and then you work the rest out later. But prioritize the three to seven things that you’re going to be doing in the year because if it’s not prioritized then everything’s a priority and if everything’s a priority, nothing is. When we kind of started executing on this mission we only had a three point plan. We were trying to condense everything down so that we could remove as much … We in fact also had a plan of what we weren’t doing as well just to make it absolutely clear because everyone loves to argue the gray and we didn’t have time for gray, so we would shut projects down and we would try and remove gray where it existed. Of course your view of where you want to be in the future has got obstacles, right? I’ve spoken about a few of the obstacles that we had. You’ve got to call those out because otherwise you’re going to be living in a dream world.
You’ve got to be explicit about what obstacles you know you’re going to need to overcome as a team and then for each method, for each thing you’re doing, we’d have a metric so we could know ‘how we can get there’. And these metrics were things that were not vanity metrics that were ways that we could chart our progress so that every month and every quarter we could sit down, honestly as a team, and see where we were doing versus where we wanted to get. I think the big clarifying thing of that between company values and a really short mission – I won’t abuse the Mark Twain quote of “didn’t have time for a short letter so I wrote a long one” – but brevity is the hardest thing in business. Brevity of your pitch, brevity of your proposition, and brevity of your plan. No one reads a hundred page power points on your plan, but they’re very easy to create. So effectively you should have a plan which is as short as a menu. Because you’ll spend ten minutes reading a menu when you’re having your dinner. You should be to spend ten minutes reading a one pager that is going to set your direction for the year. What we did is we had one page for the company and then each functional team kind of set their own plan and then we could do a better cross checking in alignment between that to make sure we didn’t have any gaps.
Making sloppy decisions
It was a really useful exercise to get a really tight plan. But while not driving everyone to micromanagement because it gave everyone freedom on how they’d actually deliver these things. that was one of the key things we learned. The other one was a bit more basic which is we just had to stop making sloppy decisions. What is a sloppy decision? Well it’s one where you haven’t really had that precision and clarity and crispness around thinking it before you start to execute on it. Again in the session earlier about entrepreneurship, one of our previous speaker spoke about how developers just want to start coding, they don’t want to talk to customers. And I’ve got a by no means exhaustive list but a few sloppy decisions that that we made that we had to stop making. And they might be familiar to you. One is we’ve got funding, we need to hire to a plan which says we should have this many salespeople this many engineers – let’s get hiring because there’s such competition in the market. We just made sloppy decisions there because we were hiring to a plan, not hiring to need, versus hiring to need you see where things are breaking and then you add resource there. We made sloppy decisions on hiring because we know we needed someone to fill that role and we’d lower the quality bar so we could get someone in there because we justify to ourselves well, it’s really hard to hire function x here, so this guy I’m sure he’ll be coached to be great. Right. And I will say the worst roles to lower the bar on are roles where you have a multiplier effect of a B or C player has an infectious effect on those around them. Couple of really good examples of that are product management and product marketing because they are a multiplier; they collaborate across a team so if you don’t have a really high bar there you’re going to have really bad product and you’re going to have really bad messaging for your sales and marketing team. the lesson I learned on that is that you cannot let the bar lower just because you have to need.
Thinking things through
Other things: we kicked off new projects where we hadn’t really thought through the overall complete business plan the business model because we will figure out as we go along – you never really figure out so do it at the start. And we kicked off projects where we didn’t really focus on the known unknowns that we need to do hypothesis testing. This is turning into a therapy session here… entrepreneurs anonymous! You can see the whole load of things that we made the icon of sloppy decisions. The bottom one is a really interesting one: over-engineering, especially when you are building a platform business, because who are your customers? Developers! Developers might be doing anything. It’s really easy to self-justify to yourself that I need to overengineered this because I’m supporting any kind of use case and you know we just had to resist temptation on that to say well we’ve got no data points that prove that there is positive demand on there, so we’re just going to pull back from kind of over-engineering.
Of course as well as working with your employees you’re working with the board. What I would say is if you’ve got a good board, your board are resilient. But what I did find is there were a few lessons that I learned along the way of how you do board communications – and they’re quite standardized and you may do them already so they might not be that insightful. This is not an actual picture of my board by the way; just if you were wondering. I would say to use the Radiohead songs by choice B for my walk on music – no alarms and no surprises. The last thing that a board member wants to do when they turn up to a board meeting is learn something… surprising right?! They need you need to go in there so they know any surprises there are. If you drop anything in a meeting you’re going to get ripped apart there. The reason for board meetings is to focus energy on strategic decisions, not just looking at the hockey sticks going up or down. What I found is that for the board meetings I’d make sure that there were no surprises, that there were close clarity over the strategic topics, and we would come out of that with a commitment on what our plan would be and what the next step of that would be. Oh, then mid-quarter I’d do a flash board update which might be as simple as an email where you could literally just read the headlines if you wanted to or read the detail it was there as well. We would tell them about progress against those strategic initiatives. We talk about where we thought we were going to end the quarter in terms of revenue. Checking in the last two weeks of the quarter is a good way for you to gently drop a little bomb if you need to about that there might be going plan or we might be exceeding it. And then you’ve got a chance to a couple of calls on those things with the board. And then end of quarter, great, where we landed, what we’re doing next. And I found that just making sure that people understood what the communication plan was, that this is how often I’ll come out to you in a formal way, meant that we got fewer inbounds from the board because they knew that I’d be coming back to them in a couple of weeks with an update on that strategic initiative where we would be and everyone kind of managed because they knew what the communication structure was. Sounds really easy doesn’t it? Of course there were surprises and there were alarms occasionally but I think on the whole you know the board really kind of understood the way, the cadence that we have with them.
And then I’d probably say then the last kind of big major one was actually just the headline.
I don’t know why but an A-Team photo in there, I just love the A-Team right and I start to think, were the A-Team volunteers? I mean how did they get paid? But maybe they were volunteers. it’s the headline that’s important thing here – treat employees as volunteers. Because you’re asking them to join you on this mission right. They’ve got choices around what they do. And that really is an ethos for a leadership style that you want. If you’d been the hard-nosed micromanagement CEO, everyone would have left because you needed to create a combined spirit that we can dig our way out of this drama, like to build kind of the technology in the business around that. That’s a mindset really of making sure we’ve got a shared mission and purpose that is a plan that comes together that we build together rather than I’ve built it all and I’m just going to dish out the tasks to each of you. It’s getting people to agree and set expectations for deliverables and they make themselves accountable for that. And you know of course I’m going to come up and say transparency is important but there’s nothing to be gained by information hiding, right. Information sharing is all positive. Okay. Because if you hide information one day people are going to find it out and you’ll be in a worse place through that. Treating employees as volunteers. I do simple things right. Not just the one on ones with my team which I did about two or three times but also shutting down my beautiful corner office and moving out to be with the rest of the team because you’ve got to remove this ‘them-and-us’ type barrier because I’m just digging my way out of this with all of you together. Which means that you’re an individual contributor as much as you are a leader. As part of that.
So my last slide. I’m gonna make it sound easy now. “Oh we did it”, right. Over a period of about 18 months we executed a really, really hard technical mission and a really hard business one; which was a pivot really. We had to build an entirely new technology for privacy by design analysis and I’m not giving lip service to privacy here because our technology was fortunate enough to have all of the posts, comments, and likes globally across Facebook kind of like streamed through it and we absolutely needed to build privacy controls that would make sure that we could, firstly, only work with anonymous data and then secondly just provide trend level insights – a bit like when you use Google Trends and it tells you kind of what direction things are going. I’m really not doing justice to the massive amount of work that we had to do to execute the pivot. But we launched the product, we built an entirely new platform business of that. Interestingly we didn’t take a bottom up seed and grow type model, we took a top down approach. We worked with large customers, strategic accounts only. We solely focused on account based marketing and account based kind of sales through that and we grew our ARR from zero to twelve million in about five quarters, which again would put you pretty high up there on the growth trajectory. We lost employees along the way but I think that what we really found out was that we came out a much stronger company at the end of that because we all felt that we’d had a shared purpose as part of that, shared values which I spoke about earlier. And it kind of created that… I don’t want to use the phrase ‘Dunkirk Spirit’, it sounds like I’m starting to get too emotional here. But you know the collective kind of mission that we had to really kind of execute came out of that.
I’m hoping you’ve all got questions for me on this. Or if you’ve not it’s because you want to get to the bar! But I do want to just go back to the first slide I’ve got. Yes, this is hard but as I said to Greg back in the day I say to you now: keep these things in perspective. We used to joke that we’re all going to at least leave this venture with all of our limbs intact. We still walked away. We built the basics back up. We were acquired at the end of that which I’m going to keep as another story for another Business of Software. But we’re fortunate to work in the industry and the dramas that we dealt with are pale by comparison to the dramas that many deal with. With that I’m going to thank you for your attention, pause, and then open up for questions.
Audience Member: Thank you, good session. one of the questions I’ve got is when you were working sort of going from Twitter and trying to work with Facebook and LinkedIn, what level we having to sort of sell into and get engagement from in terms of the business and what was the sort of time and effort that that took? One of the problems I find with those very large accounts and large customers is they’re really really hard to get the right people and to keep that sponsor for the whole duaration of what you’re trying to do – how do you do that?
Tim Barker: This would be a camera off answer for that question…
Audience Member: I have a question about exactly that… So here you are in this unexpected emergency situation. You have to switch from a peacetime CEO to a wartime CEO to quote a smart person. And you’re if you’re an engineer, you’ve got this aggressive timeline and this technical challenge. How do you avoid switching from ‘I’m going to set the vision and the goal’ to ‘I’m just gonna tell you what to do and you do it right now’?
Tim Barker: Yeah really good question and I think again there’s no simple answer to it other than the key people that they can take that message down is your direct leads; your CTO, your CPO, those people can really in some to some degree it was a question of like we will have to deliver on this but what we deliver can be flexible and the great thing about software is you can improve it tomorrow. The first thing was like get to minimum scope so you’re empowered to identify MVP and your first customer for an MVP will be an internal customer because we’ll get an internal team to go out and try and execute some projects. We’ll see if we’ve productized the data correctly. I think perhaps by being completely transparent with people like, look we’re nine months to cash out on this and no V.C. is going to put money into something that’s a bit of a restart really. You’re not going to be able to go and raise another round of funding or if you do you will definitely not like the checks that come back. I think to some degree that became the siege mentality that people needed so that we weren’t bullshitting them that we’ve got a deadline. The important word is the word dead of deadline. that really focused the minds as part of that. And we did, when I look back our first product was embarrassing, but it delivered what was needed so that we could prove the value in the data so that we can then prioritize it. In fact our sales team went to market for three months, selling a product which didn’t actually exist yet which is an impressive feat to do with PowerPoint alone. But again we just couldn’t wait for everything to synchronize as part of that. The nice thing of that was we created demand because we could say “we are coming to you to see if you want to get on the waitlist”. Now a waitlist sounds really compelling doesn’t it? Verses me saying I’ve got no product to sell you. That was probably through transparency and just through not being the micromanager on that people kind of responded to that call the call to arms.
Audience Member: My question – I’m not sure if it’s a camera off or a camera on- your original business model was 80 percent dependent on one partner and then they got you. And then your second business model was then 80 percent dependent on another partner. Yeah. So what did you do different? What did you do different in the second round to protect the company, create some more durability, sleep at night, be able to be eligible for selling it?
Tim Barker: Yeah really good question. I think the main thing was… what is it they say, often tactics drives strategy? Once we’d realized we had to build a privacy-by-design analytics platform that could work with big data and give you insight into protecting users identity, we suddenly figured that that would be a really good strategy to deal with lots of different data types. Turns out if you look in the media it probably is, there’s a lot of people that are focused on how they could combine GDPR with analytics and privacy. You know the other side to this and you know I’ll review this later on maybe – a camera on or camera off one – that to some degree, when Facebook say look we want you to partner with us, how brave a CEO would say “no, because I don’t want to get hurt again!”. So we built a technology to be data source agnostic. We proved it there by adding LinkedIn as a partner and again we had to install it behind their firewall, you know, doing privacy safe analysis there. And that’s one key reason why Meltwater has acquired us because the world’s hungry for data. But increasingly regulation is going to be there to protect it for all of us. The short answer to the question is tactics drive strategy. And also entrepreneurs get to write their own historical narrative on their successes – which is what I’ll be doing!
Audience Member: This is a selfish question on a secondary point, but as a company spread over the UK and the US with presumably most of your customers in the US (or a strong majority of them) where were you based and what was the sort of spread of your exec team and how did you manage?
Tim Barker: I was based in the UK. I spent one week a month or every six weeks in the US. The rest of my executive team, I was lucky that we actually had them scattered. We had one of our execs in our New York office. The remainder of them in our San Francisco office our CTO in Reading, and the benefit of that was whether it was good news or bad news we had local strong leadership that I totally trusted and do today to retell our message there as well. And I think the other you know if I could have added 10 or 12 different data points but clearly the strength of your executive team is another key thing as part of that you rely on. You totally can do that, engineered in Reading Berkshire, distributed in San Francisco you can totally do that. My lesson learned though was we did that in very early in our lifecycle. We opened up our San Francisco office when we launched as a company. We had people with like three or four years of tenure as part of that. They were a bit seasoned by that time for the ups and downs that we had.
Audience Member: Hi. So in terms of the capital structure. After you had the shock did you rebase manager incentives? And did you look at how to basically incentivize employees financially to stick with you?
Tim Barker: Yeah really really good question. This would be a camera off answer.
Mark Littlewood: So I have a question. You’ve been bought, you’re working happily with your acquirer. You’re obviously not going off to do something else because you’re definitely not an entrepreneur … are you?
Tim Barker: I can assure you here, I’m definitely not going to do a business that depends on the partners’ platforms. And I think probably looking back at it it’s obvious right platform risk was the biggest issue as part of this.
Mark Littlewood: Well I was going to say that your first very successful foray into that was essentially building on Salesforce platform.
Tim Barker: It was. That’s right. My first start up – I founded 3, exited 2 – the third one actually exited last year but I’d left it. So yeah I started a company that Salesforce became our first customer of and then acquired us about 2007. But I was very early in SaaS. I you know I started in the SaaS space in 2003. It was either going to be that or open source and I’m glad I chose SaaS. So yeah, I started the company and I think the lessons learned on there are very similar to the ones of DataSift. It’s all about relevance, market relevance and emotional relevance to the problem your solving. It’s all about low barriers to entry. We basically built something which became a collaboration product which is now known as Chatter.
It was very early on in kind of a whole collaboration/Slack era and yet it’s a successful product. It was a successful company. We just got taken out of the market very early. We were a seven person start up in a shop above Maidenhead, that was our office. Anyone in here from Maidenhead? Yeah, good. Shout out to the Maidenhead entrepreneur massive! On a Friday we were in Maidenhead a seven person company and Monday we were in San Francisco working in Salesforce which was a eighteen hundred person company by that point. Now it’s a 25,000 person company. And the reason I left was I thought it’s just destined to be super successful so nothing I do good or bad can stop this. So I shifted back into the world of pain and pleasure as well. Thank you for the therapy!
Learn how great companies are run
Sign up for a regular dose of actionable and useful content: