What the impact-effort matrix is
The matrix has two axes. Impact is the expected effect on the outcome the team is trying to move if the item is done. Effort is the total cost of doing it: people's time, budget, the help of other teams or suppliers, and the coordination that ties them together. Each item on a list is placed on the chart relative to the others, and its position puts it in one of four quadrants.
Its strength is speed. A list of dozens of audit findings, fixes and ideas can be sorted in a single working session, and the result is a picture that everyone in the room can read. That makes it a natural tool for the end of an audit, the start of a quarter or any moment when a long list needs to become a short plan. The guide to auditing a Google Ads account, for example, ends with a findings sheet that scores each finding on impact and effort before it becomes a backlog.
What the matrix makes visible is the gap between what matters and what is easy. Lists tend to be worked in the order they were written, or in the order that is most comfortable, and the chart shows how much of that comfortable work carries little weight. What it hides is everything that is not on the two axes: how sure the team is about each estimate, what could go wrong, which items depend on others and which have a deadline.
The four quadrants
| Quadrant | What sits there | What to do |
|---|---|---|
| Quick wins: high impact, low effort | Broken tracking, wasted spend that can be switched off, missing settings, obvious errors on key pages | Do them first, with an owner and a date, and check that each one worked |
| Major projects: high impact, high effort | Site rebuilds, new campaigns or channels, restructuring an account, new measurement | Choose a small number, scope them properly, break them into stages and test the riskiest assumption first |
| Fill-ins: low impact, low effort | Cosmetic fixes, tidying naming, small copy edits | Batch them and do them when there is spare capacity, or hand them to whoever is closest to the work |
| Money pits: low impact, high effort | Large pieces of work that would make little difference to the result | Drop or defer them, and record why so that they do not return in every review |
Quick wins are the reason the matrix is popular, and they deserve their place: work that costs little and moves the result should not wait. Two cautions apply. Some apparent quick wins are only easy because their effort has been underestimated, and some are only high-impact because nobody has checked. Each one should be verified after it is done, so the team learns whether its estimates were right.
Major projects are where the matrix is most often misused. Because they are hard, they tend to be postponed while the team clears quick wins, and the quarter ends with nothing large started. The better habit is to commit to a small number of major projects at the outset and protect capacity for them.
Estimating impact and effort honestly
Impact needs a defined target. Name the KPIA metric tied directly to an objective. that the list is meant to move and a time horizon, such as the next quarter, and estimate each item's effect on that one measure. Without a common target, impact becomes a measure of how strongly the person proposing the item feels about it. Useful evidence includes how much spend, traffic or revenue the item touches, how large the problem is, and what similar fixes achieved in the past. Write one sentence of reasoning for each estimate so that it can be challenged.
Effort should be estimated by the people who will do the work, and it should include everything: design and build, approvals, third parties, testing, measurement, and any ongoing upkeep. Items that need another team, a developer queue or a supplier are almost always bigger than they look, because waiting time counts as effort for planning purposes.
Use a simple scale, such as low, medium and high, and calibrate it. Agree a reference item for each level before placing the rest, so that "medium effort" means the same thing for everyone. Place items as a group, because disagreement is information: an item one person rates high-impact and another low usually hides an assumption worth surfacing.
Flag the items where impact is a guess. An impact estimate is a HypothesisA testable statement of the change to make, the effect expected on a named metric and the evidence-based reason for expecting it. about what will happen, and for some items the honest answer is that nobody knows. Where the item is testable and the potential impact is high, the right move may be a small test, such as an A/B TestControlled comparison of two variants., before a full commitment.
From matrix to backlog
The chart is a sorting step, not the plan. To turn it into a backlog, take the quick wins first and give each an owner, a date and a check that it worked. Choose the major projects the team can genuinely staff this period, break each into stages, and move the rest to a later review. Batch the fill-ins. Write down the decision to drop the money pits.
Then order within each group by deadlines and dependencies, which the matrix ignores. A tracking fix that every other impact estimate relies on should go first even if it looks modest, because until it is done the team cannot see the effect of anything else. A fix that must land before a seasonal peak outranks one that can wait.
Review the chart after each batch. Compare actual effort and impact with the estimates, and move items that have changed. Over time this calibration is what makes the estimates worth trusting. In a growth team the re-sort often happens in the weekly growth meeting, when a change in the north star's inputs moves an experiment up or down the backlog. The matrix also works on work the team already does: the tutorial on reviewing recurring marketing work places each report, meeting and routine on the chart to decide what to keep, shrink, change or stop. For a backlog of process and system fixes, the guide to prioritising marketing operations work uses the matrix as its first sort and weighs risk items by their consequence rather than their place on the chart.
Move to a scored model when the matrix stops discriminating: when most items crowd the middle of the chart, when the list is mostly experiments, or when stakeholders contest the order. The guide to prioritising marketing work and capacity scores requests on objective fit, impact, confidence, effort, timing and dependencies. For test backlogs, the guide to prioritising tests adds the traffic each test needs, which often decides the order on its own.
Worked example (illustrative)
A garden centre sorts its audit findings
Larchmallow Garden Centre, an invented garden centre with a web shop, commissions an audit of its paid search and website before the spring season. The audit produces 34 findings. The marketing manager, the agency and the web developer place them together on the chart in a 90-minute session, measuring impact against online revenue over the next quarter. Six findings land among the quick wins, including these: the purchase tag fires twice on the confirmation page, inflating reported conversions; 18% of spend goes on search terms such as "garden centre jobs" that need negative keywords; and four product categories send ads to a generic home page. Two are major projects: rebuilding the slow, outdated website, estimated at £45,000 and four months, and setting up a product feed for shopping ads. Nine are fill-ins, such as image alt text. Five, including a full redesign of the blog, are money pits and are dropped. The tracking fix goes first because every other estimate depends on it. The product feed is scheduled for March; the rebuild goes to the autumn plan, after a test of a faster product page template.
Limits and criticisms
Two estimates hide most of what makes prioritisation hard. Confidence, risk, dependencies and deadlines do not appear on the chart, so two items in the same spot can deserve very different treatment. A high-impact item the team is unsure about, and one backed by strong evidence, look identical.
The estimates are biased in a predictable direction. Teams tend to overrate the impact of their own ideas and underrate the effort of work that crosses team boundaries, so items drift towards the quick-win corner. Reviewing past estimates against what happened is the only reliable correction.
Quick wins crowd out the large projects that matter. The chart invites a team to clear the easy corner first, and there is always more easy work arriving. Without a deliberate decision to fund major projects, a team can spend a year on quick wins and change little.
The axes are relative. A position on one team's chart only means something compared with the other items on the same chart, so charts from different teams, or from the same team at different times, cannot be compared or combined. The matrix ranks a list; it does not measure value.
Related frameworks
When the list is mainly ideas to test, the experimentation loop and ICE scoring add confidence as a third estimate and turn the backlog into a cycle of tests and learning. The growth marketing guide shows how a backlog prioritised by impact against effort sits inside a weekly growth process. To make "impact" mean the same thing across a whole team, the KPI tree breaks the main outcome into the measures that individual pieces of work can move.
Common mistakes
- No common measure of impact. Without one named KPI and time horizon, impact means enthusiasm.
- Managers estimating effort. The people doing the work know what it takes, including the waiting.
- Only ever doing quick wins. Protect capacity for major projects before the easy corner absorbs it.
- Treating the chart as the plan. Deadlines and dependencies still decide the order within each quadrant.
- Never checking estimates. Compare what each item actually cost and achieved, or the estimates never improve.
- Comparing charts across teams. Positions are relative to one list and mean nothing outside it.
