Quarter Board · 12-week operating plan
An SEO Roadmap Your Team Can Actually Ship
Most planning documents are forty slides of ambition with no owner, no date and no definition of done. Ours is a board: four quarter columns, every task carrying a start week, an end week, one accountable name and an acceptance check you can argue with in a review.
- Diagnosis to prioritised backlog in the first three weeks
- Four quarter columns with explicit exit criteria
- One owner and one reviewer per task, written on the card
- Acceptance checks tied to URLs, not to opinions
- A decision gate every four weeks, so scope can move
The failure pattern
Why most SEO plans go quiet in month two
Month one is easy. Everyone agrees the site needs work, the audit gets circulated, the kickoff deck lands well in the leadership meeting and the agency looks organised. Month two is where the plan meets the development queue, the legal review, a product launch that outranks everything, and the uncomfortable question of who actually owns the fix.
We have inherited enough of these documents to recognise the pattern. Three failure modes account for most of them.
The plan is a list of tactics, not a sequence of decisions
Technical fixes, content, links and localisation all appear as headings. Nothing tells you which one unblocks the others. When the sprint fills up, the team removes whichever item looks least urgent, which is usually the one that had no visible consequence attached to it.
Nobody owns the second step
A recommendation to consolidate forty thin category pages is only useful if a named person writes the redirect map, a named person reviews it and a named person deploys it in a specific week. Without those three names the recommendation becomes a screenshot in a slide deck that everyone has already seen.
Success is defined as a ranking, so progress is unfalsifiable
If the only stated outcome is "rank higher", every week is simultaneously on track and off track. Nobody can accept a task as finished, so nothing is ever finished. The board grows comments instead of shipping work.
The antidote is not a better-looking deck. It is a narrower unit of work with a testable finish line.
When a card says "rewrite the eight highest-impression product pages, publish by week nine, reviewer: head of product, acceptance: each page answers the top three on-page questions in the first two hundred words", the conversation changes. Either it happened or it did not. That clarity is what keeps a programme alive past the novelty stage.
A roadmap is not a forecast. It is a set of bets with dates attached, plus the discipline to re-price those bets every four weeks.— planning principle we apply to every engagement
This page describes the structure we use, the five inputs it needs before a single card is written, and how the review cadence keeps it honest. If you want the timing picture first, our guide on How Long Does SEO Take sets expectations for each phase described below.
Anatomy
What a shippable SEO roadmap contains
Five components. Remove any one and the plan degrades into a wish list within a quarter.
1. A diagnosis with evidence attached
The diagnosis is a short document, not a two-hundred-slide audit. It states the commercial constraint, the three biggest reasons the site is not earning the traffic it should, and the evidence for each. Evidence means crawl exports, query-to-page maps, log samples, revenue attribution by template, competitor visibility share. If a claim cannot be traced to an export, it does not enter the roadmap.
2. Prioritisation that survives a budget meeting
We score each candidate workstream on four axes: expected impact on qualified organic sessions, confidence in the mechanism, effort in engineering days, and reversibility. Reversibility matters more than most teams admit — a redirect map is cheap to reverse, a domain migration is not. The score is not a magic number. It exists so that when a stakeholder asks why content comes before links, the answer is already written down.
3. Quarter-level decomposition
Each quarter gets a theme, a small number of workstreams and an explicit non-goal. The non-goal is the part teams skip and later regret. Writing "we are not touching international expansion this quarter" prevents three weeks of speculative meetings and keeps the delivery team on the actual bottleneck.
4. Owners and acceptance criteria
Every card names one accountable person, one reviewer and one acceptance test. The test is written in the past tense at planning time: "the page is live, returns 200, is included in the sitemap, is linked from the hub, and the primary query has a dedicated answer block". That single convention removes most end-of-quarter disputes about whether work was done.
5. Review gates that force a decision
Every four weeks the board is re-priced. Three questions: what shipped, what the data says about it, what gets cut to protect the next commitment. A gate without a documented cut is not a gate, it is a status update.
The artefact
The quarter board, in four columns
The board is deliberately plain. Four columns, one per quarter, each with a theme at the top and no more than six cards. If a column needs twelve cards, the quarter is over-committed and the plan is lying about capacity.
Reading the board in five minutes
Start at the top left. The first column answers "why are we not already winning"; the second answers "what is structurally in the way"; the third answers "how do we create demand we can capture"; the fourth answers "how do we compound instead of restart". A new stakeholder can follow that narrative without a meeting, which is the point.
Milestones are decisions, not decorations
The diamonds on the board are not celebration markers. Each one is a decision that must be made on a specific date: rebuild or consolidate, hire or outsource, scale the cluster that worked or abandon it. If a milestone passes without a written decision, the following quarter inherits unresolved scope, and that is how programmes quietly double in cost.
How it gets built
Five passes, roughly three weeks
- Constraint interviewNinety minutes with the person who owns the revenue number. We establish what a qualified inquiry is worth, which markets matter, and what the business cannot change this year. Everything downstream is filtered through that conversation.
- Evidence sweepSearch Console, analytics, crawl exports, server logs, CRM source data, competitor visibility. We look for the gap between pages that earn impressions and pages that convert, because that gap usually contains the cheapest available win.
- Candidate generationEvery plausible workstream is written as a one-line card with an estimated effort band, then scored. We deliberately over-generate here so the shortlist is a choice rather than a default.
- Sequencing workshopYour delivery lead, our strategist and whoever controls the development calendar sit with the shortlist and build the four columns together. Capacity is the constraint, not ambition. This is the session where a roadmap becomes yours rather than ours.
- Handover and first gateThe board goes live in your project tool with owners assigned, plus a scheduled review four weeks out. The first gate is the real test: did the cadence survive contact with a normal working month?
Timing
The first ninety days, plotted
Nothing dramatic happens in weeks one to four, and we say so up front. That period is spent making the site legible: fixing indexation, removing duplicate templates, establishing a baseline you can trust. Weeks five to twelve are where template-level changes start moving impressions, because a single well-built category template can lift hundreds of URLs at once.
Content compounds later. A cluster published in week ten typically starts earning stable positions in weeks twenty to thirty, which is why the board reserves Q4 for refresh work rather than hoping new pages win immediately. Teams that expect month-one traffic from month-one content end up cancelling in month three, right before the curve turns.
The practical rule: if a plan cannot point to one structural change and one content change that will both be measurable within ninety days, it is not a plan yet. It is an audit with dates.
Evidence
Three boards, three shapes
An industrial components supplier, 6 months: organic inquiries 5→23/mo after a category template rebuild and forty-page consolidation. The first quarter shipped almost no new content.
A B2B software vendor, 9 months: non-brand organic sessions 3.1× with the same headcount, because the board reserved two weeks per quarter for refresh work instead of new pages.
An exporter with four language sites: crawl spend on duplicate parameter URLs cut by 38%, freeing indexation for the pages that actually carry product intent.
Engagement
Scope before price, always
We quote after the constraint interview and the evidence sweep, because a number quoted before those two conversations is a guess dressed as a proposal. Scoped work typically covers the diagnosis, the board build, a monthly recalibration and a written gate memo each cycle. Larger programmes add cluster production, digital PR and technical delivery support.
If you already run an in-house team and only need the planning layer, we can deliver the board and coach your lead through two cycles. That model works well when the bottleneck is decision-making rather than execution. Most clients move to an ongoing arrangement once the first quarter proves the cadence, which is why we keep a Monthly SEO Retainer structure available rather than forcing a long annual contract.
Bring us the plan that stalled
Send the roadmap you already have. We will mark up which cards are unowned, which acceptance checks cannot be tested, and where the sequence breaks — before you commit to another quarter of the same document.
Before the workshop
Questions we get in the first call
- How long before a roadmap produces results?
- The board starts producing shipped work in week two. Measurable movement in impressions and click-through usually appears between weeks eight and sixteen, and it depends mostly on how much of the backlog is template-level rather than content-level. Template work moves fast because one change touches hundreds of URLs.
- Is the roadmap a fixed twelve-month document?
- No. We commit to one quarter in detail, sketch the next two and treat the fourth as a direction. Each gate re-orders the backlog using evidence from the previous four weeks, so the document stays a working tool instead of a museum piece.
- Who owns the tasks, us or you?
- Every card has exactly one accountable name. Content production and technical fixes usually sit with your team because they touch your stack and your voice. Research, briefs, internal link maps, outreach and reporting sit with us. The split is visible before work starts, which prevents the classic month-two standoff.
- What if our developers cannot take the work?
- Then we rewrite the board around what can ship without them: consolidation, internal linking, on-page improvements, schema and content refresh. A plan that ignores delivery capacity is fiction, so capacity is an input in the sequencing workshop rather than a surprise in week six.
- How do you define done?
- The URL is live, returns a clean status, is internally linked, appears in the sitemap, is measurable in analytics and has been accepted by the named reviewer. A ticket marked complete without those checks stays on the board until it passes.
- Do we need a monthly retainer to keep it running?
- Most clients continue monthly because somebody has to run the gate and re-price the backlog. The first quarter can also be delivered as a standalone sprint if you want to test the working style before committing further.
- Can you work with our existing agency?
- Yes, and it is often the fastest route. We take the planning and measurement layer, they keep execution, and the board becomes the shared contract between both teams. Ambiguity about who does what is the single most common cause of stalled programmes.