Scale Well
To grow at speed, first lay solid foundations. Here are 6 steps:
1. Get to P/M fit first.
Premature scaling is the number one cause of startup failure. Scaling your cost base without a market to support it is a fast track to nowhere. How do you know you have product-market fit? Here are some thoughts.
2. Get good, then get big (aka best beats first).
So you’ve got P/M fit? Good for you. Don’t kid yourself, you aren’t ready for world domination yet. Now is the time to start training. You need to get fit before you can run the race. Start by getting good at reaching one segment of customers, with one brand proposition, with one product. Figure out what type of people work best for building/selling/supporting/marketing your core product. You will need to iterate on these things and get them tuned to a point where you have a recipe for exploding your market share. You should be able to give this recipe to a team who don’t know the product and they should be able to succeed without you. Test your recipe book in a new geography / language / adjacency. If it works, then your training is done.
3. Sort out your knowledge sharing.
When you’re 5 people in a room there is ample opportunity for everyone to be in each other’s head. That has to stop. Start early — every time someone learns something new make them share it — wikis are your friend. Asynchronous knowledge sharing = leverage. You will need all the leverage you can get. Don’t let the training in point 2 go to waste. Codify it and make it easily available to your team. This will enable to you get more out of each new employee you hire, more quickly — that’s leverage. Speaking of new employees…
4. Hire well. Hire right.
So you’ve got good and you’ve documented how to do it. It’s almost time to put the pedal to the metal. There is just one more thing: getting the right type of people. You should have figured the characteristics of the people who can make your product succeed in step 2. Now you need to find more of them. Adopt a structured hiring process. Be explicit about the qualities you are testing for and how your hiring manager should test for them.
5. Over-communicate.
Nature abhors a vacuum. So do your employees. As your org grows it is easy to lose track of who knows what. Get into the habit early of running regular all hands, Q&As and/or updates. If it’s not your thing, then give someone the responsibility of making sure everyone is across the context and objectives of the organisation.
6. Decentralise decision making.
To move with speed, you need to move as many decisions to the extremities of the organisation as possible. This is made possible by steps 1–5. Be deliberate and explicit about which decisions you keep centrally controlled and which you push out. Note that decentralisation will eventually cause a crisis of control and bloat your organisation with duplication of work. You won’t believe me at the time, but this is a good problem to have. Fix it as your focus shifts from growth to profitability.
Give me six hours to chop down a tree, and I will spend the first four sharpening the axe. — Abraham Lincoln
A case study: Early Uber vs Early Lyft
Lyft launched its low-cost ride-hailing product 12 months before Uber, yet today Uber is in ~900 cities across a dozen countries and Lyft is in ~300 cities in two countries. Both scaled, but one did better. Let’s explore why:
Get to P/M fit first.
Both Uber and Lyft got to P/M fit. It was easy enough for both of them to know when they got there. Both had to control user signups to maintain reliability in their early days. When you struggle to maintain a reliable supply of your product, you probably have P/M fit. ✔
Get good, then get big (aka best beats first)
Uber, however, had the opportunity to get good at launching. UberBLACK had launched in 7 cities before Lyft launched in SF. UberBLACK was in 24 cities before UberX started. Uber had spent 3 years getting its playbook together before it really hit the scale button. Whilst Lyft had the jump on the low-cost product, Uber had been training for 2.5 years and was ready to run.
Sort out your knowledge sharing.
Uber became adept at storing, sharing and refining its corporate knowledge. The most far-flung launch teams had access to the sum total of institutional knowledge via the company’s internal wiki: team.uber.com. If a solution was not forthcoming in the wiki, it would at least point to the right subject matter expert, who would inevitably be available and willing to work through an issue day or night. Such was the culture; credibility was earned by contribution to others via wikis or by being available to spread knowledge. As a result, every launch team had a better playbook than the last.
Hire well. Hire right.
From early hiring successes and failures, Uber determined the types of people that worked well for scaling the company. Whilst this would eventually sow the seeds of behavioural issues once the challenger became the incumbent, for early Uber this cookie-cutter approach of hiring hyper-competitive hustlers worked well. As the ‘good guys’ Lyft tended towards hiring ‘good people’ as opposed to win-at-all-cost types. The lesson: hire for qualities that are logically coherent with your strategy and business context.
Over-communicate.
Hour-long townhalls with open Q&As are the staples of most tech scale-ups. However, it is perhaps telling that Uber ran its all hands weekly, while Lyft did theirs fortnightly.
Decentralise decision making.
This is perhaps the biggest differentiator between Uber and Lyft. Uber made clear, deliberate judgements as to which decisions would be pushed to the edge of the company and which would be centralised. Uber pushed as many decisions to the edge as possible, holding on only to key branding, tech and product suite decisions. This lead to lightspeed decision making. Lyft, anecdotally, was plagued by indecision and disagreement among key executives. This hesitation resulted in the forfeiture of key international markets to Uber.
Use the 6 steps to project yourself throughout your business
The combination of taking the time to figure out what the right thing to do was, codifying it, making it accessible, hiring for it and updating it constantly via scaled communication gave Travis Kalanick the superpower of being in multiple places at once. He converted himself from a founder into an operating system.
A note on post-blitzscaling
After successfully scaling, Uber experienced many more challenges. It has had to walk away from unprofitable markets in Asia and Europe. It has had to endure battles with regulators. It has had internal behavioural issues. The seeds of these failures were sown in the company’s success. It pays to remember — the solution to a crisis sows the seeds of the next crisis. Knowing this gives you the chance to see around corners and mitigate the next crisis.
Get Fletcher Young’s stories in your inbox
Join Medium for free to get updates from this writer.
Few lessons from post-scaling mistakes:
- Being a challenger is different from being the incumbent, know when you’ve made the switch.
- Scaling a business is different from running a business at scale. They require different skill sets. The boundary between the two is difficult to pick at the time but obvious in hindsight.
- The people and culture that were an asset during scaling may become a liability once you have a dominant market position.
Working through the 6 steps
You can be working on steps 2–6 concurrently, but you will want to complete them, approximately, sequentially. There is a lot more to each step than is contained here. There are threads of strategy and culture that run through the 6 steps, so it is crucial to have a good handle on these.
If you want to assess your readiness for scaling, get in touch: https://au.linkedin.com/in/fletcheryoung
