At Reforge, we’ve written about how companies actually grow, and built an entire program around it. Most companies, when they talk about how they grow, will usually pick from one of the following terms:
- Sales Driven e.g. Oracle, Workday
- Marketing Driven e.g. Hubspot, Moz
- Product Driven e.g. Atlassian, Github
- Engineering Driven e.g. Google, Palantir
- Product + Sales Driven e.g. Slack, Stripe
- Marketing + Product Driven e.g Uber, Amazon
- Sales + Marketing Driven e.g. Drift, Salesforce
Most people can’t reasonably answer why they are one of these types, but there are reasons. If people inside these companies have thought enough about it, they might understand the market to have attributes that force these styles:
- Sales Driven: Custom value props and big customers
- Marketing Driven: Need to make a “space” that doesn’t exist yet and convince people they need it
- Product Driven: High viral quotients and/or word of mouth from product differentiation
- Engineering Driven: Solving hard technical challenges creates markets
- Product + Sales Driven: Large customer spread with bottoms up adoption
- Marketing + Product Driven: Marketing fuels network effects
- Sales + Marketing Driven: Custom value props and big customers and need to make a “space”
That was extremely simplistic, but hopefully you get the idea. The larger a company grows, the more likely singular definitions like this start to break down though. Companies launch multiple product lines that require different distribution models, and these product lines typically build on top of each other that gain from the intersection of them. In the Advanced Strategy course, we teach how to model these systems of loops that power such companies.
What growth teams sometimes miss is that optimization is not always the answer to a growth problem. It may require a new product or building a sales team. Modeling your loops to understand your constraints to pick tactics that alleviate those constraints takes some time to do well. A framework I’ve used for more quick and dirty decision making is using data companies usually have on hand: whether a tactic is reliable or not historically in driving growth i.e. you can predict the output from an investment or not, and how fast the payback period is.
For those who aren’t familiar, payback period is one of the most important metrics you can track for growth. It’s very simple. Given this investment today, how long will it take before I recoup that investment in profit from the customers it impacts. For example, a customer you acquire in Adwords for $10 might take you six months to make $10 in profit, subtracting all marginal costs. So our payback period would be six months.
Most executive teams can start to plot where in this 2×2 tactics seem to sit. For example, brand marketing at all but the most sophisticated marketing driven companies is something that has a slow payback and is not reliable. Similarly, innovative product development focused on entirely new products or value props usually sits in that category. For other tactics, where they sit in this 2×2 may vary. On Pinterest’s growth team, for a more specific example, efforts in product driven growth around UGC content distributed through SEO, conversion optimization, email marketing, and activation were very predictable and had quick payback periods, but viral growth was unreliable. At Uber, I imagine viral growth via incentivized referrals was very reliable, but SEO was not. Now, it’s important to remember that within reliability is a sense of scale that matters for the business. If a tactic gives you .1% growth, and only 10% improvements matter, it actually isn’t a reliable lever. An alternative is to make a three dimensional chart where reliability is separate from impact, and ain’t nobody got time for that.
In the long run, model your loops well and find the constraints. In the short term run, maximize efforts on reliable and quick payback activities until you hit diminishing returns. Then, think about moving excess resources into things that are reliable, but have longer payback periods. And think about how anything with short paybacks that are unreliable can become more reliable.
What you’ll quickly realize in the case where you have built a proper growth model or you’re short term optimizing based on this 2×2 above is the opportunities are rarely siloed to one function. You can’t even build this 2×2 if you don’t have many functions represented. So start talking to all the functions of your company to map the opportunities to grow better so that you can grow faster.
Currently listening to Simplicity is the Ultimate Sophistication by Matthieu Faubourg.