A Framework for Scaling product teams

In the Invention phase, the company is an idea trying to prove that it can exist. The company doesn’t know what signals to pay attention to. An Inventing company will change directions quickly as it hones in on what the company will be. The people need to be flexible. The processes need to be light touch to reduce change management and wasted effort building processes that won’t last. The systems must rapidly evolve and adapt to the changing processes.

In the Craft stage, the company has a direction and the growth will come from nailing that direction. A Crafting company will stay in one lane, driving toward a mostly-understood target. The people need to be creative and collaborate well. Processes need to enable people to have ownership and execute decisions quickly. Systems need to support creativity over compliance. A Crafting company doesn’t care if they’re doing everything the best way, they care that they’re rapidly improving.

The Optimizing stage is marked by stacking many small improvements. Growth comes from the compounding effect of hundreds of single-digit percentage improvements that last forever. The people need to specialize and have redundancies. The processes need to be repeatable and more standardized so that an improvement can improve every team at once. The systems need to support knowing and tracking minute details and rolling those up into broad KPIs.

These stages are company stages, not product stages. A company that’s reached the Optimizing stage will still introduce new products. However, the company will reject the efforts of the teams that build those products the same ways an earlier-stage company does. A company must build products using methods that match its stage.

Inventing

An Inventing company should only hire product people when it’s incredibly painful not to. Because things are changing rapidly, the need you identify today could be gone tomorrow. To avoid this, wait until the need has been around for a while. New hires should bring something completely new to the team. When hiring a second designer, look for someone with research or illustration skills if your current designer is more UX-focused. Or bring on a product manager with enterprise experience if your team today is mostly from consumer backgrounds. Everyone hired into a product role should be comfortable with ambiguity and unstructured environments.

Teams in Invent stage companies are largely autonomous, with every team free to set their own processes. If coordination between teams is needed, it’s ad hoc. Once the organization grows to multiple teams that need to roll up information, some lightweight company-wide processes emerge. This emergence should happen because someone else needs something they aren’t getting and that need is re-occurring. For example, if marketing keeps needing a unified roadmap, then it’s time to create just enough process to maintain a cross-team roadmap.

At all stages, the company is driving toward a long-term vision. However, the horizon a company pays close attention to is shorter at each earlier stage. At this Invention stage, the product plan may only stretch a few weeks. The frequent pivots and reactions to early customers mean longer plans are pointless.

The purpose of software systems is to support process. Since there’s little process, there’s little need for systems. Wikis and spreadsheets can go a long way and are easier to throw away or radically change as the company figures out how to work.

Systems at the Invent stage don’t need standardization across teams. If part of the company wants to use Miro and part wants to use FigJam, that’s OK. Systems only converge where it’s necessary to support sharing across teams.

Crafting

A Crafting company starts to add structure and layers to the organization. Scaling the organization’s skills starts to come from training and less from hiring unique skills. More specialization and more redundancy are needed. The hires must be high-agency. In the Invent stage, comfort with ambiguity was a key skill you looked for. In the Craft stage, you’re looking for a different sort of grit. You need entrepreneurial types that can identify the right problems to attack even when they have incomplete data.

Hiring should still come later than you think it should, but particular attention should be paid to creating new roles to grow the leverage of the existing team. A dedicated researcher or data scientist who speeds up your existing product managers can help with scale more than hiring 5 new PMs.

The company’s main need at the Craft stage is to ship and create value quickly. Process should free the product organization, not control it. The purpose of process at this stage is to create clarity about constraints and boundaries. People should know what decisions they can make and have a framework for getting quick support on decisions outside their scope.

You’ll also need to create processes to support the new, more complex structure of the Product organization and the company. It’s getting harder for every product team to know what every other one is doing. Other parts of the company need more information and certainty from the product organization. Changes in direction become more disruptive. Communication slows down. To scale at this stage you’ll need stable, longer-term product plans.

The complexity, need for more robust communication, and desire to control costs will drive Craft stage companies to converge systems. It’s time to standardize most things. It’s also time to add systems to support fast decision making. This includes product analytics and data science tools that let product managers learn, find patterns, and measure the value they’re creating.

The product team will also start building support systems to help the company scale. Tooling for support. Capabilities that allow customers to self-manage.

Optimizing

When Optimizing, the product organization specializes and focuses on organization structure. The company’s growth is coming from doing everything a little bit better every day. Product management leans heavily into experimentation. The main skill you hire for is finding the best solution to a known problem. The people who thrived in ambiguous and creative stages may struggle at the Optimize stage, and you’ll need strategies to retain or replace them.

The product organization has predictable attrition, so you should create a steady hiring plan at this stage. You build a recruiting pipeline to backfill an anticipated number of openings each quarter.

Processes at this stage are designed to create certainty. Your product is mature enough now that customers, partners, and the market at large will expect more certainty from your roadmap. You get this certainty with a much heavier process touch, trading speed for precision. Product plans look a quarter or more ahead.

It’s likely that you have a process for everything and many teams have little freedom to decide how they operate. The organization is large enough that you’ll need to tolerate lower-performing teams, but you’ll drive those teams through process.

To scale at the Optimize stage, you need systems to support experiments and research. A team can’t experiment their way to a great product, but they can use experiments to find slight improvements.

Your goal with systems at this stage should trend toward enforcing compliance over enabling creativity. You achieve certainty through compliance. The only way to optimize everything is to have everything operating in similar manners. You find efficiencies through repetition, and your systems should help you create this process of repetition.