Building is fast now, so product has to move faster too. Without a clear strategy, you find out that:
- product managers have to settle trade-offs that should already have been settled. Which customer comes first? Do we build this for one big client, or keep it generic? Do we charge for it, or give it away to grow? Each question means a meeting, a wait or a guess, so product managers aren't ready when development is;
- different product managers go in different directions. If nobody settled the trade-offs, each one settles them their own way, and fast building turns a small difference in direction into an inconsistent product;
- there is no time left for course corrections. The weeks of planning that used to absorb a wrong turn are gone. By the time you notice the direction was wrong, it is already built.
Product can only run fast if it knows where to run to. It can't spend the marathon figuring out which course to follow. That course is the product strategy: written down, owned by one person, and put where product people and their agents read it.
What a product strategy is
Many engineers, and even product people, have never worked with a product strategy. It is not a roadmap and not a list of features. It describes the product you want to be in one, two or three years, and the areas you therefore have to invest in.
- Know what a product strategy is: what you will do, what you won't, and why, but not how. It says who you serve, which problem you solve for them, what you deliberately don't do, and how you choose when two good things compete. What a feature looks like is decided later, while scoping. It forces you to make a picture of what you want to be, so you don't spend all your time on the small features existing customers ask for, without making the big steps your ambition requires. A strategy solves four problems:
- the big steps get made, not only the small requests;
- product managers answer the same questions the same way, without a meeting;
- a feature that doesn't move the needle gets stopped before it is built, not after;
- the people and agents who scope an item can check it against the strategy on their own.
Example, not a real company
Where we want to be in two years
The tool small agencies run all their client work on, from proposal to
invoice, not only the invoicing.
Where we invest
Proposals and contracts clients sign online, time tracking that turns into
invoices without retyping, and reporting on which clients are profitable.
Trade-offs
- Agencies of 5 to 50 people, not freelancers and not large consultancies.
- Adoption before monetization: we don't charge for what gets people started.
- Self-service over professional services: if it needs a consultant, we
simplify it instead.
- Configuration over bespoke features: one product for every customer,
never a version for one.
Writing it
A strategy only helps if it answers the questions product managers will run into, and nobody can predict all of them. So it has to settle the principles, not every case, and someone has to keep it clear.
- Write the key trade-offs into the strategy. Trying to answer every question in advance is big design up front, one level higher. State the trade-offs clearly enough that product managers can answer the questions you didn't predict, and answer them the same way. Don't try to answer feature questions this early: you only know what you want once you see it running, so what a feature looks like and does gets decided while scoping.
- Make one person responsible for the product strategy. Without an owner, every product manager works out the direction alone, and each ends up somewhere different. Make one person responsible, usually the Head of Product. They keep it clear, and they answer the questions product managers would otherwise have to figure out themselves.
Using it
A strategy in a slide deck helps nobody at the moment an item gets scoped. It has to be where the work happens, and something has to hold every item up against it.
- Put the product strategy in a file your agents read. An agent only knows what it can read. If the strategy lives in a slide deck or in somebody's head, the agent that helps a product manager scope a feature works without it. Put it where your agents can read it, especially the agents your product team uses.
- Make the scoping skill check every item against the strategy before it is scoped further. With items scoped many times a day, the one who holds each item up against the strategy can't always be a person. Have the scoping skill read the strategy first, and say plainly when an item doesn't fit, before anyone spends time scoping it. The product manager still decides; the skill makes sure the question is asked. How scoping works is on the Scoping page.