How to build a growth roadmap

What a growth roadmap is, and what it is not

A growth roadmap is the answer to "what do we do next, and in what order?" It lists a small number of projects for a fixed period, each one linked to a goal and a number you will check.

It is not a content calendar, a list of tools or a year-long plan. Markets, search results and customer behaviour shift too much for that. A short roadmap that you update beats a long one that nobody trusts. The person who owns it is usually the growth manager, described in our guide What is a growth manager?.

Step 1: set one goal you can measure

Start with the result that pays: booked consultations, qualified enquiries, orders or revenue. Traffic and followers are inputs, not goals. Write the goal as a number with a date, for example "40 qualified enquiries a month by the end of the quarter".

Google’s re:Work guide on OKRs (objectives and key results) gives a useful discipline. It recommends three to five objectives and around three key results for each, and says the sweet spot for achieving them is a score of 60 to 70 percent. If you always hit 100 percent, the goals are probably too easy.

For a small business, one objective and three key results is enough. Pick numbers your marketing KPIs can actually track.

Step 2: find the constraint

The theory of constraints, developed by Eliyahu Goldratt, says every system has a constraint that limits its output, and improving anything else does not raise profit much. The TOC Institute lists five steps: identify the constraint, exploit it, subordinate everything else to it, elevate it, and check whether a new constraint has appeared.

In growth terms, lay out the journey: visit, enquiry, qualified lead, sale, repeat. Calculate the conversion rate between each stage. The lowest or most broken stage is usually the constraint. If 2,000 people visit a month and 10 enquire, adding traffic is the expensive way to grow. Fixing the page that loses them is the cheaper one.

Constraints move. Once enquiries rise, the limit may become how fast you answer them, or capacity to serve. Re-check at each review.

Step 3: list ideas aimed at the constraint

Collect every idea from the owner, sales team, support inbox, analytics and reports. Then keep only those that act on the constraint or that unblock it. Everything else goes on a parked list with a date to look at it again.

Write each idea as an action and an expected effect: "Shorten the quote form from 11 fields to 5, to raise the share of visitors who finish it." That wording makes the idea testable.

Step 4: score with ICE or RICE

ICE is credited to Sean Ellis, who coined the term growth hacking. According to ProductPlan, each idea gets a score from 1 to 10 for Impact (how much it moves the key metric), Confidence (how sure you are) and Ease (how little effort it takes), and the three are multiplied. In ProductPlan’s example, an idea scoring 7, 6 and 5 totals 210 and beats one scoring 9, 7 and 2 at 126, because low ease pulls the second down.

RICE adds Reach. Sean McBride at Intercom described it in 2018: Reach is how many people are affected in a period, Impact is scored 3 (massive), 2 (high), 1 (medium), 0.5 (low) or 0.25 (minimal), Confidence is 100, 80 or 50 percent, and Effort is person-months. The score is Reach times Impact times Confidence, divided by Effort.

Use ICE when speed matters and you have little data. Use RICE when you can count reach, because it stops a small idea that touches few people from outranking a big one. Treat both as a way to force a conversation, not as a source of truth. The inputs are judgements, and different people will score differently.

Step 5: sequence and plan 90 days

Sort by score, then adjust for dependencies. You cannot test a new offer page before the tracking works, and you cannot measure leads without an agreed definition of a qualified lead. Put these enabling jobs first even when they score low.

Google’s own cadence is annual and quarterly OKRs, with check-ins at the midpoint and end of each quarter. A 90-day cycle suits most small businesses for the same reason: long enough for work to show up in the numbers, short enough to correct course.

Limit the cycle to three to five projects. If something is added, something comes off.

Step 6: review on a fixed rhythm

• Weekly, 15 minutes: check leading indicators such as form completions and calls. Fix anything broken.
• Monthly, 60 minutes: compare results with the goal, mark each project as continue, change or stop, and decide the next month’s order.
• End of cycle: score the cycle against the goal, record what you learned and write the next roadmap.
• Always record the reason behind a decision, so that the next review can test it.

A worked example

An online shop sells to 20,000 visitors a quarter. Its goal is to raise orders. Mapping the journey shows that most people who reach the cart leave before paying, so the constraint is checkout, not traffic.

It lists three ideas and scores them with RICE. Fix the mobile checkout: reach 4,000 cart visitors, impact 2, confidence 80 percent, effort 1 person-month, giving 4,000 x 2 x 0.8 / 1 = 6,400. Send an abandoned-cart email: reach 1,500, impact 1, confidence 80 percent, effort 0.5, giving 2,400. Start a blog series: reach 10,000, impact 0.5, confidence 50 percent, effort 3, giving about 833.

The order is clear: checkout first, abandoned-cart email second, blog third or parked. The numbers are illustrative, but the reasoning is real: the biggest score comes from acting on the constraint.