Surviving the messy middle.

“Can I tell the truth? If I was doing this for you then I’d have nothing left to prove. This for me though, I’m just trying to stay alive and take care of my people”

2019 was an interesting year for us, and in retrospect the best thing ever to happen to us as a company, and aside from my son being born, maybe me as a person.

While there were ups and downs in years past, we had always delivered, growing almost 100% year over year the previous few years. Then we hit the messiest part of the messy middle. Like really hit it. There is no shame or ego here, and this certainly isn’t another post about failure, and all the things we should have done. For what it’s worth, I hate those posts, they make me cringe. This is a post about how we got through it and it laid the foundation for sustainable success. My hope is that others can read this post, and enjoy smoother sailing.

While companies with big expectations are assumed to be crushing and killing it, I never felt that pressure. Admittedly I don’t subscribe to most of startup garbage advice that you read on Medium (the irony). But last year we hit an important moment along our trajectory. I love the image below from Scott Belsky’s book, appropriately titled “The Messy Middle” because it so adequately represents the complete journey, the ups and downs.

If I had to guess, we were squarely in the “WTF happened?!” zone in the middle of 2019.

The start.

When we launched mParticle in 2013 we had a vision of what a modern data platform should look like. We didn’t call it that at the time (we called it a Data Automation Platform), but it was based on our experience at our previous company building our flagship data platform called Open Segment Manager (OSM), which afforded us a growth trajectory strong enough to take the company public and then grow into the market leader at the time of our acquisition by Yahoo in 2011.

We architected mParticle to address modern business needs around successful user-level data consumption, while contemplating privacy by design from day one. We launched the company as a tool for marketers to be able to move faster, and eliminate engineering dependencies. The early days of mParticle weren’t pretty, it took us longer to get traction than anticipated and we almost ran out of money. We also almost got acquired three times in two years and then we didn’t, and then we found product market fit and didn’t look back.

We championed the problems around “SDK fatigue” and as mobile became the primary screen we found product market fit. There were ups and downs but the net result was significant growth, working with the biggest brands in the world, and three successful rounds of investment in about the same amount of time it took us to get our initial traction. Our go to market centered on the end consumers of data (marketers and analysts), and over the next couple years most of our own marketing investment was on big events and experiences in order to evangelize both the brand, as well as the category. We put on some pretty incredible events.

It worked, until it didn’t.

WTF Happened?!

They say that companies begin to struggle not because they do things that don’t work but because they continue to do the things that did work but they do them for too long; I wish someone told me this a few years ago.

Heading in 2018, I’m not sure we saw a single RFP or RFI for a Customer Data Platform; but by the end of the first week of January 2018, we received more RFPs than we did the previous 3 years. 2018 was largely a successful year for us, at least on the surface. We grew nearly 100% and did some exciting deals with some great companies, but we failed to recognize a massive shift was coming. Sometime in 2018, the Customer Data Platform category had started to evolve, new entrants in the space emerged, and the market began to shift. The category became noisy but we felt confident in our strategy and roadmap, working with some of the largest brands in the world, solving foundational customer data challenges. So we convinced ourselves that no changes were needed; there’s a fine line between confidence and arrogance. But something felt off…

Lesson #1: If it feels early, the timing is probably right. Don’t wait.

As we got into 2019, we believed that we could continue to grow nearly 100% again but by mid-Q1 realized that the CDP market as we knew it, had completely shifted. Hindsight being 20/20, we should have focused on solving customer problems and reinvested marketing spend into R&D. Instead we continued to invest in go to market tactics that had worked quite well for us in years past but were becoming largely inefficient. We incorrectly assumed that we could muscle our way through it and ignored the diminishing returns, chalking it up to delayed feedback cycles in marketing an enterprise software product.

What we didn’t realize heading into 2019 was that as the CDP category was forming, the buyer was changing. Almost everyone thinks about the CDP as a tool for marketers, its not. It’s infrastructure to enable customer engagement at scale (of which marketing is an important component) by solving a number of foundational data challenges. As the market recognized and demanded infrastructure to address perpetual fragmentation, the buyer persona had moved from data consumers to data providers; from Marketing and Analytics groups to Product and Engineering groups.

By the end of Q2 2019, we knew that things needed to change. The thing about momentum in a startup is that it works in both directions, and unfortunately for us we had momentum taking us in the wrong direction. We started the process to course correct by first addressing the culture, attempting to heal from within. By the end of the summer in 2019, we rolled out our new values and implemented a V2MOM process to create a new cultural operating system. By the late Q3 2019, we had seen enough and knew real change had to be made.

Lesson #2: Don’t try to muscle or market your way through turbulence. Retreat and refocus, stay close to the customer.

The ascent to YES!

The end of Q3 2019 was the first time since the early days of mParticle where I struggled to see how it was all going to work out. I had made a number of mistakes along the way, and confidence was wavering across the organization. We went from convincing ourselves that we were crushing it, to having to deal with reality head on. This meant making significant changes. I wasn’t sure what that meant but what I did know was that as CEO if I didn’t “take my seat”, as my executive coach Jerry Colonna likes to say, that our fate would be sealed.

To be honest, I didn’t know what my plan was going to be but I knew that we had an incredibly talented team, and as CEO it was on me to make sure we figured it out. I recognize that my title does not entitle me to anything, and that the outcome may have very well been failure but it was up to me to at least clean up my mess, because as CEO this was my mess. Again, I share the lessons I learned from this process, not to glorify the struggle but the hope that I can help others avoid the mistakes that I made.

During the process, I told the team repeatedly that, to quote Ben Horowitz “There are no silver bullets, only lead bullets.” I made this my mantra, and hearing the quote to this day puts a smile on my face. We had to put in the work, and there wasn’t going to be a quick fix or a single initiative that would fix everything for us. In order for us to get back on track, we needed to focus, and re-prioritize a number of important initiatives; and we needed to work our asses off.

Part of that work for me meant fixing our go to market. The market changed, and our GTM needed to change as well, and unfortunately we couldn’t rely on anyone else to know what we needed to do. While we had a good head of sales who helped us grow, a change was needed and we mutually agreed that the market evolved to a point where it made sense to part ways amicably. Along the way, we had also hired too many sales reps and let the marketing team get bloated. We parted ways with our Head of Marketing shortly thereafter as well.

Previously, we avoided making organizational changes in fear of what kind of message that would send to the market, but we no longer had that luxury. The only things that mattered were the things that mattered. In all we ended up re-orging about 10% of the company, all in Sales and Marketing. To the people that didn’t make it along for the ride, I personally apologize for the missteps along the way and for the temporary turbulence we created for you both personally and professionally. I do truly hope everyone has found great opportunities that give them a greater opportunity for success than we did.

In some ways, this change was liberating. I remember distinctly in November of 2019 walking outside and feeling a sense of calm, and having complete mental clarity. I knew what we needed to do, and was relieved that we weren’t distracted by ego, past success, or the stories we told ourselves. We just had to do the fucking work, and pretty much nothing else mattered. It was the happiest I had been in awhile. And as for worrying about the market, nobody cares — for better or worse, nobody cares what you have to do.

Lesson #3: The market doesn’t care what you have to do inside your organization to fix stuff. Don’t worry about what the market may think, just get it done.

YES!

Free of constraints and misalignment, we immediately implemented a number of changes that I will discuss shortly. It felt a bit chaotic at first, I would imagine somewhat like a home renovation project. We dove in and uncovered a number of incredibly valuable insights that were hiding in plain sight, and thanks to the incredible work of people throughout the company we were able to respond with our best quarter ever!

Q4 2019 was the first quarter that the new sales system was put in place. It produced the most amount of new revenue in a quarter that we had ever achieved, along with the most number of new customers that we had ever won in a quarter, and we did it with half the number of sellers, and no marketing investment.

We followed that up with an incredibly strong start to Q1 2020. We ended up putting up our best first quarter we ever had by about 40%, and then just as we had started to really regain momentum and confidence was growing, COVID hit. Having successfully operated through the financial crisis in 2008, COVID was both different and unprecedented.

We went fully remote in March, and cut whatever remaining marketing investment we had allocated. What followed was an emergency board meeting, scenario planning, and a conservative forecast for the rest of 2020. I remember thinking “just when things were starting to go well, we were hit with a global pandemic, this is it.”

Uh Oh!

As an entrepreneur, operating with uncertainty becomes normal after a while but as we started Q2 2020, we had to come to grips that this was going to be anything but normal. Unemployment had spiked to over 20M people in April. A number of publicly traded companies issued revised and more conservative forecasts. We had a number of customer opportunities start to delay, opportunities that were supposed to close imminently. The team was starting to feel fatigued from an unprecedented number of Zooms per day, the quarantine was making everyone stir crazy, plus social injustice and racism was only getting worse. I think everyone reading this can relate, it’s still fresh in our minds and very much unresolved.

We knew that we were going to have to operate differently. This started with how we work, one of the best changes we made was pushing for a writing-first culture. While we admittedly have a long way to go to to do it more consistently, this made a big difference in terms of alleviating *some* Zoom fatigue, but more so getting people to communicate and collaborate remotely. The value of creating these collaborative artifacts has been insanely valuable as proper context and decision logs. The other area was addressing issues that were always taboo in the workplace; directly addressing the issues around politics, social issues, and everything else that was weighing heavy on everyone’s mind. Addressing the elephant in the room is about living our values, welcoming everyone’s views and perspectives, and giving everyone a voice.

The other change that we made which I believe is a game changer was to implement a day each week where there are no internal meetings. I used to think that this was overkill, and that as adults we can all appropriately manage our calendars but this is about alignment. And while meetings inevitably creep onto the calendar, its the best day of each week and by far the most productive.

Lesson 4: Improving the quality of our daily work environment is way more important than the work itself.

$$$

While there was turbulence for some customers, the nice thing about selling infrastructure to large companies is that they typically have the means and wherewithal to withstand most turbulence. We also saw a number of very promising signals. We had started to see a significant increase in activity from a number of our existing customers, and began to see inbound lead volume begin to accelerate even though we were literally not spending a dollar on marketing.

In all, Q2 2020 ended up being our most successful quarter ever from a new revenue perspective. Not only were we able to do more than we had ever done before, we were able to do it with less. By the end of the first half of 2020, we had closed 50% more new revenue than we did in the first half of 2019, and Sales & Marketing spend had decreased by 50%; with marketing programs decreasing by about 80%.

We followed that with our best Q3 ever by a wide margin, beating our internal targets by about 70% and growing 70% from what we were able to accomplish in Q3 2019. At the midway point through Q4 2020, we are on our way to achieve a new high watermark, with almost no marketing investment (yet).

The story is still playing out and if we are on a trajectory that’s consistent with the chart I started the post with then we will hit another set of challenges soon. So while I am sure there will be ups and downs along the way, I think we have proven its all part of the journey.

My takeaways:

As CEO, I’ve always been pretty hands on with GTM both on sales and marketing, I find that it helps me keep pace with the market, and have exciting conversations about what our customers are doing, and where there may be opportunity. I recognize that I’m the annoying founder who wants to evaluate our messaging strategy, so we go through a multi-month messaging exercise and by the time it’s done I’m unhappy because it’s already out of date.

So as I took over sales, I have to admit I was nervous. I had never run sales before, nor had any desire to. I knew what I didn’t like and what didn’t work but that didn’t necessarily mean I knew how to fix anything. I had always outsourced the Head of Sales function to someone else since really early on, and never spent the time necessary to learn the role first. In turn, the biggest mistake I made was not trying to truly understand all of the mechanics first before bringing in someone better than me.

My first bit of advice before I start for any founder is that if you are going to eventually bring in a head of sales (and you should), you should first spend the time understanding the drivers of your sales process first. It’s different at each company and be weary of anyone with a “playbook.”

Learning what’s needed for successful sales and go to market execution at each stage will also teach you what you need in a sales leader, when the time comes. There are lots of different types of sales leaders: the big deal type, the relationship type, the process type, the manage everything from a spreadsheet type, the big company (at scale) type, the scrappy early stage type, the sales lead from another industry who doesn’t get the space but has a great resume type…it’s dizzying and it’s super easy to choose the wrong type of sales lead.

After being on the job for a year now, what I’ve come to appreciate is that sales is a system, and requires systems thinking. It’s not about process, or a playbook, or methodology, or relationships, or the funnel, or enablement. It’s about all of it, mapping the inputs and outputs, aligning and optimizing how it’s all connected via information flows. And while one of a startups greatest assets in the early days is that it can do things that don’t scale, if you want to create deterministic outcomes then at some point that has to change.

Lesson 5: The faster you can move the business from analog to digital, systematizing every function the bigger and better the company will become.

The tl;dr is that managing sales at scale is about system design. How that system can be improved by leveraging data to create the necessary feedback loops and fed back into the business which improves the output over time. I’ve learned a lot, and am still learning a ton but here are my observations so far:

1. There is no such thing as a playbook for success. Every company is different but businesses are systems, and the systems are a number of interconnected subsystems. Certain playbooks help but they are highly dependent on the context. Are you customers b2b brands or b2c brands? Are you up market or down market, is it an established category or an emerging category, are you selling to technical or non-technical stakeholders? Is it a land and expand motion or are you looking to maximize economics on day 1. Is it a high velocity, transactional sales process, or is it a consultative motion with large deal sizes.

To founders, stop thinking that someone else has all the answers. We all have fears but outsourcing those fears doesn’t fix them. Internally, we talk a lot about the physics of our business, which is to say the way that our business moves. Applying a “playbook” that worked somewhere else is no better than guessing what may work. Typically it’s the sales leaders from big companies who come in with their “playbook”. The person at the big company with a great resume may seem attractive on paper but chances are they’re a cultural mismatch, they won’t know how to sell your product, and they may be great when the hard problems are already figured out but they usually can’t help you before that.

2. Communication is about alignment, not celebration. Most sales leaders will focus just on the sales team, after all managing the sales team is what they were hired for. They’ll set up a recurring sales meeting where everyone on the team meets, and talks about their deals in front of everyone else. The problem is that everyone just waits for their turn to speak and nobody else listens as you go around the room or video conference. Nobody learns anything and then time expires and everyone goes back to doing what they’re doing. Total waste of time. The weekly sales meeting has to include everyone who contributes to revenue. Cofounders, Product and Eng leads, Marketing, Sales Enablement, Solutions Engineering, and Customer Success.

Yes it means the meeting gets bigger but it becomes more useful. The weekly sales meeting isn’t a role call or a celebration. It’s about creating a dialogue so that the sales team can get everything they need to be successful, and that requires key stakeholders from around the organization.

3. The journey of a thousand miles begins with one step. That means you aren’t going to fix everything all at once and you have to embrace continual improvement. We started by making sure the daily and habits and behaviors are there for monthly and quarterly success. One thing we did is that we started tracking everyone’s opportunity creation activity. Not to force everyone to send a bunch of cold emails, in fact cold outreach should be the last resort. But it’s about focusing a certain amount of time every day to building a pipeline, whether it’s through inbound, outbound, events, partnerships, referrals, social; it doesn’t matter. We wanted to create a system of accountability by showing a leaderboard to the team and to the company.

4.Play to learn, then to win. Look, no seller likes using Salesforce, and I don’t blame them. It’s boring software with an awful interface, but it’s so important. Use a tool like Troops, it helps.

When I took over the sales team what I found was that the information being produced was utterly trash. We had six different reasons we were using to explain why we weren’t winning deals, and zero reasons to attribute to a success. Of the six “Closed Lost” reasons, four of them were all saying the same thing just different ways — the deal wasn’t qualified. Utterly useless…

The data entered into your CRM needs to be complete enough to answer difficult questions such as why am I winning or why am I losing deals, so that ultimately the other parts of the org can provide the support required to make sales successful.

5. Chop it up! The data that is captured needs to be cut a number of different ways. We weren’t doing any real sales analysis prior but I suspected if we get more granular we could uncover some interesting trends and insights. We started to analyze the data through a number of different lenses. Sure enough, as we started to look at wins and losses by sales rep, by buying persona, by market segment, and geography we saw some clear trends. For example, we saw that we had really high win rates in certain areas and really low win rates in others. That allowed us to ask the necessary follow up questions to invest in areas.

6.Build the system. It’s way too easy for a sales leader to find themselves under the gun, especially if they’re not thinking about building a sales system. If this happens, they may go into self-preservation mode which means they aren’t trying to keep the job, they’re just trying to keep it for as long as possible. So what happens next? You guessed it. They start playing games with data that goes into Salesforce to make the pipeline look stronger than it is. They may push for selling more to existing customers, or introducing service sales; all of which mask the real underlying issue.

This is a high level schematic of our sales system. The goal of the system is to capture market demand (created by marketing, partnerships, sales people, dev rel etc), transform it into opportunities, and then convert it to sales/revenue. Along the way, information is created which needs to be incorporated into the business. Each of these functions runs on their own system as well, for example there is a system for Product development and input from the market is one data point but there is much more complexity to it.

Hope this was helpful to someone. Peace.