Don’t kill your startup by forgetting about the gear-shift from lab to factory
Things that brought your early-stage startup to product-market-fit will probably kill you if you keep doing them in scale mode. I think about it as a gear shift, that needs to happen before you start scaling. For me things started breaking the first time I got >1m in VC-funding and grew the team to over 10–15 people. I just kept on “managing” the company as I always did, because I didn’t know any better and because I never saw how management should look like since I never had a real job before. That led to a lot of frustration among employees. People quit, customers got angry and I didn’t know what was going on.
Don’t scale things that don’t scale
I look at an early stage startup as an experimental lab to try out things and figure out what works. A scale-up however should look more like a buzzing factory where the machines and production-lines should work in harmony.
In an early-stage startup it’s necessary to do things that don’t scale. For example bring on users manually and talk to them in person, use manual processes that humans (instead of computers) do in the beginning and so on. But please don’t try to scale those things once you hit product-market fit.
That was probably one of the bigger mistakes I made. Just because something was working in “experimental mode” did not mean it was scalable. For example the setup-process for new customers at my company RegioHelden was completely manual in the beginning. After asales rep would bring in a contract someone would manually create a Google AdWords account, enter the keywords and onboard the customer. That was perfectly fine to prove that such an offering could work and create value for a customer that they would pay for. We could nicely calculate unit economics like customer-acquisition-costs and customer-lifetime-value. But of course it wasn’t really scalable.
I always thought that once I got the VC-funding I’d just hire a bunch of developers to automate the processes and make them scalable. After all, it wasn’t rocket science. We set up AdWords accounts, so how hard could it be?
In the end it was true. We hired the developers, we automated the processes and now everything works more smoothly. There was just one problem: It took much, much longer than I thought. I’m talking about years instead of weeks or months like I thought first.
And that brought us into a tricky situation: Of course I used the VC money to scale the sales and marketing processes to achieve the growth I promised the investors. So we increased the amount of new customers by 5x but didn’t have any automated processes in place. We were both in experimental-mode and in scale-mode at the same time.
That’s a bad place to be in because things start breaking. The capacity of the people working in the processes isn’t enough suddenly. So we hired more. That meant more inexperienced people working on half-baked processes. Some of them got frustrated and left. So we established a recruiting department to bring in even more people. That created problems that distracted us from the actual automation we needed to build. You get the point… lots of operational complexities and not enough real progress.
What I learned from all of this: Hire the developer. Automate the process first and then scale it. Promised growth to your investors? Yeah, but they already invested anyway and if the growth comes half a year or a year later but is sustainable instead of fragile they will be happier after all.
Shift management gears
Another thing to switch is management style. In an early-stage startup you don’t need much management. Think about the experimental lab again. The people there wear lab-coats and not suits — that’s no coincidence.
Under 10 people communication is informal, everyone sits in the same room and most of the statt know (at least roughly) what the others are working on. Things like 1:1s, staff meetings, performance evaluations just aren’t that important.
..of course the communication gets complicated
Big difference in scale-mode. For me when we hit over 10 people things got messy. New people kept coming in and I didn’t train them personally. Which meant the message that I wanted to bring across got diluted. For some reason people also seemed to lose the urge to question things and figure out solutions by themselves. It was the first time I heard an explanation like “…because we have always done it this way” which scared me.
Also, my straightforward communication style (which worked very well in the early days) sometimes would intimidate employees because they tried to interpret hidden messages, that weren’t there.
I didn’t really realize there was a problem until people started quitting. They gave reason like “the need more stability” or want to have “less responsibilities in bigger companies” or “learn new things”. But it was probably the lack of proper management, that drove them out.
Founder turns/hires manager
I think people join for the company and the mission but quit because of bad management. So my biggest lesson from this phase is that in scale-mode “no management” is not acceptable.
That means as a founder you have to do one of two things: 1) learn how to manage and/or 2) hire a manager. I started with 1), then realized I’m not really good at it and I don’t really enjoy it much. So I switched to 2), hired a COO and started to let him manage big parts of the business. Way more effective, both for me and the staff.
That brings me to one of the most important questions in scale-mode:
Does an entrepreneur have to be a good manager?
I think the answer is no. And it’s also natural, that the skills you need to start a company and go through the experimental lab-mode are very different from what you need in scale-mode. Some typical founder skills are just the opposite of what you need in scale mode. Think about intensity, curiosity, lack of patience, overly ambition and creativity vs. stability, patience, routines, calmness and reliability for example.
You just have to know your own skills and act accordingly by either learning how to manage and become good at it or hire people who are better at it. In lots of successful startups there is a beautiful symbiosis between founder skills and management skills. Prominent examples are he Google founders and Eric Schmidt, or the Zuckerberg/Sandberg combo.
Be aware of the gear-shift that will happen once you switch from experimental-mode to scale-mode. It needs to happen if you want it or not. Think about a new way to look at processes as well as management style. It’s better to shift gears deliberately without overheating the engine first.