Product strategy.

Key terms

Product vision

The product’s purpose: the ultimate reason it exists and the positive change it should bring about. It changes little over the years.

Product strategy

The approach chosen to reach the vision and make the product succeed: which problems, for whom, and how you’ll win.

Product roadmap

How the strategy will be carried out, ideally as a sequence of outcomes or goals rather than a list of features.

Guiding policy

In Rumelt’s kernel, the overall approach chosen to deal with the challenge. It rules some things in, and many things out.

Positioning

In April Dunford’s words, how your product is a leader at delivering something a well-defined set of customers cares a lot about.

Monetization model

How the product makes money: who pays, for what, and how often.

Take rate

In a marketplace, the share of the money flowing through it that the marketplace keeps as revenue.

Kernel of strategy

Richard Rumelt’s name for the three parts every good strategy contains: a diagnosis, a guiding policy, and coherent actions.

The idea

“Strategy” may be the most overused word in business. Plans get called strategies. So do goals, slogans, and lists of projects. None of them is one.

This lesson gives you a working definition, a test you can apply to any strategy document, and the pieces a product strategy usually needs: focus, positioning, how the product makes money, a price, and a way to decide whether to go after a new market. The made-up furniture store runs through all of it.

The lesson, in seven parts

1. 1Vision, strategy, roadmap: how they fit.

   These three words get mixed up all the time. Roman Pichler, who writes and teaches about product strategy, puts them in a hierarchy. The product vision “describes the product’s purpose, the ultimate reason for creating it, and the positive change it should bring about.” It sets the direction you steer by (not the North Star metric from lesson 4), and it should change little over the product’s life. The product strategy “communicates the approach chosen to realise the vision and to make the product successful”; for a digital product, he says, it often looks 12 to 24 months ahead. The roadmap describes how the strategy will be carried out, ideally as outcomes or goals. Below it sits the backlog, the detailed work.

   Marty Cagan of the Silicon Valley Product Group sums up product strategy as one question: how do we make the product vision a reality, while meeting the needs of the company as we go? In practice, he writes, it decides which problems the product teams should solve.

   Layer	Answers	In the made-up furniture store
   Vision	Why does this product exist? Where is it going?	“Anyone can furnish a home they love, without the guesswork.”
   Strategy	Which problems, for whom, and how will we win?	Win first-apartment movers by taking the worry out of size and delivery
   Roadmap	Which outcomes, in what order?	Now: fewer returns for fit. Next: fewer missed deliveries. Later: more repeat purchases
   Backlog	What exactly do we build next?	Fit Check (the made-up store’s tool that tells you whether a sofa will fit through your door and in your room) on sofa pages; doorway measurements saved in the app

   Layers follow Roman Pichler’s product strategy model. The examples are made up.

   Notice what each layer leaves to the one below. The vision doesn’t pick customers. The strategy doesn’t list features. If your “strategy” is a list of features with dates, it’s a roadmap, or a backlog, wearing the wrong name. The roadmap row uses Now, Next, Later from lesson 8. There you wrote each item as a problem; here each is the outcome of solving it.

2. 2What makes a strategy good: Rumelt’s kernel.

   Richard Rumelt, professor emeritus of strategy at UCLA Anderson, wrote a book product people often cite: Good Strategy/Bad Strategy (2011). His central idea is that a good strategy has a kernel of three parts:

   • A diagnosis: what’s really going on? It names the challenge honestly and explores what makes it hard.
   • A guiding policy: the overall approach chosen to overcome the obstacles the diagnosis found.
   • Coherent actions: coordinated steps that carry out the policy and support each other.

   He’s just as clear about what bad strategy looks like. The four hallmarks he describes:

   • Fluff: a restatement of the obvious, dressed up in buzzwords.
   • Failure to face the challenge: if the problem isn’t named, there’s nothing to check the strategy against.
   • Mistaking goals for strategy: “grow revenue 30%” says what you want, not how you’ll get it.
   • Bad strategic objectives: too many, muddled together, or impossible to act on.

3. 3Write the strategy, and make the hard choices.

   Cagan says product strategy depends on four things most companies find hard: making tough choices about what really matters, finding insights in data and in customers, turning those insights into action, and managing the work actively without micromanaging. “Choices means focus,” he writes: deciding the few things you really need to do, and therefore all the things you won’t.

   Here’s the made-up store’s strategy, written as a kernel. Use the template for your own:

   The template

   Diagnosis: [what’s really going on, and the one challenge that matters most]. Guiding policy: [the approach we’ll take, and what it rules out]. Coherent actions: [three to five steps that support each other, and how we’ll know they’re working].

   Filled in, for the made-up furniture store

   Diagnosis: we sell well to people who already know what they want, but first-apartment movers, our fastest-growing group of visitors, buy less often and return more, mostly because they can’t tell whether a sofa will fit or when it will arrive. Big competitors win on range and price, and we can’t outspend them. Guiding policy: win first-apartment movers by taking the worry out of size and delivery, even if that means a smaller range. Not now: a marketplace for other sellers, a loyalty program, or chasing the lowest price. Coherent actions: Fit Check on every sofa page; a range of sofas that come apart for narrow doors; delivery windows we can keep, with tracking; support staff trained on small spaces; and success measured by fewer returns for fit and more repeat purchases. (All made up.)

   Read the actions again. Each one makes the others stronger. The narrow-door range means Fit Check says “yes” more often, and both make the delivery promise easier to keep. That’s the coherence the kernel asks for. A list of unrelated good ideas doesn’t have it.

4. 4Positioning: why customers should choose you.

   April Dunford, a positioning consultant and the author of Obviously Awesome, defines it like this: “Positioning defines how your product is a leader at delivering something that a well-defined set of customers cares a lot about.” It isn’t a tagline or a brand story. She compares it to the opening scene of a movie: it sets the context, so customers understand what your product is and why it matters.

   She breaks it into five parts, and says to start with competitive alternatives, because what makes you different only means something next to what customers would otherwise do:

   Component	The question	In the made-up furniture store
   Competitive alternatives	What would customers do if we didn’t exist?	Buy from a big online store and hope it fits, or shop in person
   Unique attributes	What do we have that the alternatives don’t?	Fit Check; sofas that come apart for narrow doors; tracked delivery windows
   Value	So what, for customers?	Buy with confidence the first time, and skip a costly return
   Best-fit customers	Who cares most about that value?	First-apartment movers in city buildings with narrow doors and stairs
   Market category	What context makes the value obvious?	Furniture for small city spaces

   Components and order follow April Dunford’s positioning method. The examples are made up.

   “Do nothing” counts as an alternative, too. Dunford says the most common mistake is in how teams define competitive alternatives. The real alternative is often the customer’s current way of coping, like a tape measure and a guess, not the competitor you worry about most.

5. 5How the product makes money.

   A strategy has to say how the product pays for itself. These are the common models. Many products mix several:

   Model	Who pays, for what	A made-up furniture store version
   Transactional	The customer pays for each purchase	Selling sofas, one order at a time (what the store does today)
   Subscription	The customer pays a recurring fee for ongoing access or service	A yearly membership with free delivery and design help
   Freemium	The basic version is free; customers pay for a premium version or extras	A free room planner; paid planning with a designer
   Advertising	Advertisers pay to reach the product’s users, who often pay nothing	Brands pay for sponsored spots in search results
   Marketplace commission	Sellers pay a share of each sale: the take rate	Independent makers sell on the site; the store keeps a share
   Services and add-ons	The customer pays for extras around the main purchase	Assembly, room-of-choice delivery, old-sofa removal

   A made-up example. Many products mix several models.

   Two things to watch. First, each model changes who you’re serving. With advertising, advertisers are customers too, and you’ll have to balance what they want with what users want. Second, in a marketplace, don’t confuse the money flowing through it with revenue. Andreessen Horowitz, a venture capital firm, points out in its guide to marketplace metrics that gross merchandise volume, the total sales on the platform, is not revenue: revenue is the portion the marketplace takes. A $1,000 sofa sold by an independent maker at a 15% commission is $1,000 of sales, but $150 of revenue for the store (made-up numbers).

6. 6Pricing: three common ways to set a price.

   Once you know the model, you need a price. The Australian Government’s business guide describes the common approaches plainly, and they apply anywhere. It lists others too, like premium and penetration pricing; these three are the ones you’ll use most to reason about a price:

   Approach	How it works	What it can miss
   Cost-plus	Add up the full cost of making and supplying it, then add a markup	What customers would pay, and what rivals charge
   Going rate (competition-based)	Price at or near the market leader; use competitors as a guide, but don’t copy them exactly	What makes you different, and your own costs
   Value-based	Base the price on the value to customers, not just the cost; test what they’ll pay	Your costs and rivals’ prices, if you look at value alone; and it needs real evidence of value, not hope

   Approaches follow business.gov.au, “Choose a pricing strategy”. The last column is general advice.

   In practice, you look at all three. A useful way to combine them: cost sets the floor, value to the customer sets the ceiling, and competitors show where customers will compare you. Here’s a worked example, with made-up numbers, for an assembly service. Staff time and travel cost the store about $60 per sofa, so cost-plus at a 50% markup says $90. Competitors charge $79 to $129. In interviews, shoppers said assembly saves them about two hours and a lot of frustration, and on a test page many still chose it at $99. So $99 sits above the cost floor, inside the market range, and the test page suggests many shoppers think it’s worth at least that.

7. 7Should we enter a new market?

   Sooner or later someone asks: should we sell outdoor furniture? Should we launch in Canada? There’s no single official method for this, but the questions are fairly standard. Work through them in order, and write down an answer to each:

   • Fit: does it serve our vision and our guiding policy, or pull against them?
   • The market: how many customers, how much they spend, and whether it’s growing. Rough numbers are fine; lesson 13 shows how to estimate them.
   • Customers: who exactly, what’s their problem, and what do they use today?
   • Competitors: who’s there already, and what would make us the better choice? (That’s positioning again.)
   • Strengths: what can we reuse, like our delivery network, our brand, or Fit Check?
   • Money: what will it cost, how will it make money, and when might it pay back?
   • Risks: legal, operational, and what happens if it doesn’t work.

   Then decide, and say how you’d enter: build it yourselves, partner with someone, or buy a company that’s already there. For the made-up store, balcony furniture for city apartments might fit: small spaces, the same customers, the same fit worries. Launching in a new country might not yet: a new delivery network, new rules, new competitors. The point isn’t the answer. It’s that the answer follows from the strategy.

With AI: AI can produce a polished strategy document in seconds, and it will often be what Rumelt calls fluff: the obvious, restated in impressive words. Use it differently. Ask it to challenge your diagnosis, to list the alternatives customers actually use, or to point out actions that don’t support your guiding policy. AI is also changing what your strategy has to face: as building gets cheaper, features may get easier for competitors to copy. Clear choices about which customers to serve, and what to say no to, tend to be harder to copy.

PRODUCT:

VISION (why it exists, the change it brings; one line):

DIAGNOSIS
  What's really going on:
  The one challenge that matters most:

GUIDING POLICY
  Our approach:
  What it rules out (not now):

COHERENT ACTIONS (3 to 5, each supporting the others)
  1.
  2.
  3.
  How we'll know it's working (outcomes):

POSITIONING
  Competitive alternatives (incl. "do nothing"):
  Unique attributes:
  Value for customers:
  Best-fit customers:
  Market category:

HOW IT MAKES MONEY
  Model (transactional, subscription, freemium, ads, marketplace, add-ons):
  Who pays, for what:
  Price: cost floor ___ | competitors ___ | value to customer ___

NEW MARKET? (fit, size, customers, competitors, strengths, money, risks)
  Decision, and how we'd enter (build, partner, buy):

ROADMAP: NOW / NEXT / LATER (outcomes, not features)

Ask AI about this lesson

All lessons