The annual marketing plan is dead.
Not metaphorically dead. Not "under increasing pressure." Dead in the practical sense that it stops working for most businesses around six weeks into Q1, and everyone in the room knows it but carries on adding rows to the spreadsheet anyway.
In 2026, the 90 day sprint has replaced it for any business serious about search visibility and AI discovery. Not because it is a fashionable idea from a business book. Because it is a realistic response to a marketing environment that is now changing faster than any twelve-month plan can track.
If you have ever written a careful annual marketing plan in January and found yourself describing it as "directional guidance" by March, you know exactly what this feels like. The plan was not bad. The world just moved.
This piece covers what a 90 day sprint actually is, why annual planning specifically struggles in search and AI marketing right now, the maths that make a quarterly rhythm work for a small team, what a real sprint looks like in practice, which metrics belong in it, where most sprints fail, and how to run your first one. It is honest about the parts most pieces skip.
What a 90 day sprint actually is
A 90 day sprint is a marketing operating rhythm. Three goals are agreed for the next quarter. Each goal has a limited number of projects underneath it. Monthly checkpoints test whether the work is moving in the right direction. Weekly stand-ups keep pace. The quarter ends with a structured review and a decision about what the next sprint is.
That is the whole structure.
The term gets used interchangeably with several things it is not, and the difference matters when you are trying to run one.
It is not the same as EOS or Traction quarterly rocks. Rocks are business-wide goal-setting tools designed to create alignment across the whole organisation. A marketing 90 day sprint is specifically about shipping marketing work. Tighter scope, different purpose.
It is not a Scrum sprint. Software development teams run two-week Scrum cycles with daily stand-ups, velocity tracking, and user stories. Importing that vocabulary into a marketing programme usually produces more ceremony than output. The 90 day sprint is deliberately looser.
It is not an OKR. The objective and key results system, originally from Intel and made famous at Google, describes where you want to go and how you will measure arrival. OKRs set the destination. The 90 day sprint describes the work you will do to get there. Both can coexist, but they are not interchangeable.
A 90 day sprint is about cadence and commitment. You pick a small number of things you genuinely believe you can ship and measure in twelve weeks, and you hold yourself to that.
Simple. Not easy.
Why annual marketing plans collapse in 2026
They always drifted. The 2026 version just drifts faster and further.
Digital marketing is now changing at a pace that makes twelve-month planning feel like guesswork dressed in a Gantt chart. Google's own core update record shows multiple major ranking changes each year, each capable of reordering visibility for entire site categories. The rollout of AI Overviews changed how a significant portion of the results page looks for informational queries. The llms.txt convention for AI crawler guidance did not exist eighteen months ago.
And that is just search. AI search is rewriting the discovery channel for whole categories of research and buying intent. ChatGPT, Claude and Perplexity now handle volumes of queries that would previously have landed directly on Google. Their behaviour, the sources they favour, and how they handle recommendation queries shifts month to month. A January plan that assumed a stable Google SERP as the primary discovery channel looks quite different by September.
Paid social CPMs have been volatile for two years. Server-side tracking has moved from optional to necessary as consent requirements tighten. GA4 continues to receive updates that shift how you interpret the numbers underneath you.
The honest read: a detailed marketing plan written in January is half wrong by April, and most teams pretend otherwise because admitting you need to replan feels like failure. It is not. It is the correct response to an environment that actually changed.
Annual plans still have a role. Direction, investment prioritisation, rough resource allocation. But as operating instructions, they are too slow for the medium they are supposed to manage. The SEO programme you commit to in January looks different once Q1's algorithm updates have shown you what actually moved.
The plan points north. The sprint moves.
The maths of the 90 day sprint
Three goals per sprint. Not five.
If you start with five goals, you will finish with partial progress on five things and nothing properly shipped. Three goals with real projects underneath them is already a full quarter for a small marketing team. And small marketing teams, meaning one to five people running SEO, content, paid and web alongside existing client work, are what most growing businesses actually have.
Each goal supports at most three projects. A goal like "rebuild authority on the priority keyword cluster" might have three projects under it: update the ten highest-value pages, publish eight pieces of supporting content, and close the technical debt backlog from the audit. That is a realistic quarter of work for one person if they are not doing much else.
[image generation failed: id=maths]
Twelve weeks is the headline. Nine working weeks is the honest version. Remove bank holidays, planned leave, the two weeks most quarters where everything disappears into reporting cycles or sales presentations, and the unplanned days that evaporate into urgent work. You have nine to ten weeks of real marketing output time.
Four planned meetings in the quarter: sprint kickoff at 60 minutes, two monthly checkpoints at 45 minutes each, end-of-sprint review at 90 minutes. Not a programme of workshops.
The arithmetic that makes this concrete: one significant piece of work shipped per fortnight is a credible output rate for a team running marketing alongside everything else. Commit to that and a three-goal sprint is full but achievable. Promise more and it becomes a list of good intentions nobody reviews.
UK businesses are already wired to a quarterly reporting rhythm whether they plan for it or not. The quarterly VAT return cycle means most businesses are reviewing their financial picture four times a year. Aligning marketing sprint reviews to that same rhythm means the "are we getting a return on this?" conversation happens when the finance data is already on the table.
A worked 90 day sprint for a UK SEO and AI search programme
This is what a first 90 day sprint looks like when the priority is search visibility and AI discovery.
Sprint goal one: close the foundational SEO gaps. Complete the technical audit and act on the findings. Rebuild internal links across the ten highest-value pages. Update the priority pages that are underperforming relative to their keyword targets. Not new content. Existing pages, made more likely to rank.
Sprint goal two: publish six pieces targeting the priority topic cluster. Not six blog posts for the sake of it. Six pieces mapped to specific search queries with genuine intent, with rank tracking live on each target keyword before publication. Month two is when most of the writing ships. Month one is planning, keyword finalisation, and two pilot pieces to test the format and cadence.
Sprint goal three: establish the AI search visibility baseline. This is the step most marketing plans skip. It means tracking how ChatGPT, Claude and Perplexity currently answer the eight to ten queries most relevant to the business, documenting whether the business is cited or absent, publishing llms.txt if it does not exist, and completing the named entity work so AI models have accurate, consistent information to work with. This is AI search visibility work and the first pass takes three to four weeks to do properly.
The monthly rhythm: month one is foundations and baselines. Technical fixes ship, tracking is live, the first content pieces draft. Month two is output and momentum. Publishing cadence holds, AI citation monitoring runs weekly, the internal link rebuild completes. Month three is measurement and sprint planning. You read what actually moved, form a view on what worked and what did not, and plan sprint two with real evidence rather than the assumptions sprint one started with.
Weekly cadence: a 30-minute stand-up on Tuesday morning, a 15-minute metrics check on Friday afternoon. The Friday check is not a meeting. It is one person opening Search Console, Ahrefs and the AI citation log and writing two sentences about what they saw.
That is the sprint. Contained, measurable, and possible to actually do.
The metrics that belong in a sprint and the ones that do not
A 90 day window is too short to wait on revenue. The lagging indicators, total sales, marketing-qualified leads converting to sales conversations, customer acquisition cost, move on a longer arc than one quarter. Expecting them to shift in twelve weeks is how you abandon work that was actually working.
The leading indicators are what belong in a sprint review.
Worth tracking: indexed page count (are the new pages being crawled?), average position on the tracked keyword set (are the priority pages climbing?), AI citation appearance rate (is the business being referenced in relevant ChatGPT and Perplexity answers?), organic landing page sessions on the target pages, and lead magnet downloads as a proxy for qualified interest.
Track but do not panic about mid-sprint: organic revenue, MQLs to SQLs, customer acquisition cost. These matter enormously. They just need more than one quarter to show meaningful movement.
Drop entirely from sprint reviews: total pageviews, average session duration, social impressions. Total pageviews rewards volume over quality. Average session duration is easily distorted by one or two pages with unusually high dwell time. Social impressions measure reach, not intent.
The tools: GA4 for landing page session data, Search Console for average position and index coverage, Ahrefs for rank tracking on the priority keyword set, Looker Studio to pull it into one view. Four tools, opened weekly, cover almost everything a sprint review needs.
A sprint review tests whether the work moved the leading indicators in the predicted direction. The annual plan tests whether the leading indicators eventually pulled the lagging ones. Both questions matter. They just belong to different timescales.
Running the end of sprint review
The review has a standing agenda. It is not a celebration and it is not a post-mortem. It is a decision meeting.
What shipped. List the work that completed. Be precise about what partially shipped. "Draft complete but not published" is not shipped.
What slipped. Every sprint has work that did not happen. The question is whether it slipped because it was harder than expected, because something more important came up, or because it was never realistic to begin with. The answer tells you something about the next sprint.
What changed in the market. Anything significant in search, AI, or the competitive landscape during the quarter. An algorithm update, a competitor launching a new content programme, a shift in how a key query is being answered in AI platforms. This question keeps the review from becoming entirely internal.
What we learned. The most important item on the agenda, and the one most often skimmed. What do you know now that you did not know twelve weeks ago?
What we commit to next. The next sprint starts here, not two weeks after the review.
The hardest question on the agenda: which goal would you drop if you could only keep two? It sounds confrontational. It is actually the question that forces honesty about what matters and what was on the list because nobody wanted to argue against it at the kickoff.
The review is 90 minutes. If it needs a full day, the sprint was overscoped. Book it before the sprint starts.
Where 90 day sprints break
Sprint theatre is the most common failure. The team announces it is running sprints. Reviews happen on schedule. Nothing changes because nobody reads the review document or acts on it. The structure exists. The decisions do not.
Goal inflation is close behind. Three goals become five, then eight, as everyone adds "just one more thing" in the planning session. Hard rule: if it is not committed at the sprint kickoff, it is not in the sprint. Mid-sprint additions are what kill pace.
The "we will plan it next week" trap. Sprints that bleed into the following quarter because the planning session was never booked. Q2 ends and Q3 starts three weeks later because everyone is still processing the review. Sprint planning must be the last agenda item in the end-of-sprint review, not a separate meeting to schedule at some point before the quarter actually starts.
Wrong cadence. Some work genuinely does not fit a 90 day window. A complete brand identity rebuild, an enterprise procurement cycle, a full website redevelopment. The 90 day sprint is a marketing operating rhythm, not a universal project management tool. Forcing work that needs twelve months into a 90 day box produces a different kind of plan that will not be followed.
Honestly, this is hard. The posts that make it sound like a five-step morning routine are usually trying to sell you a Notion template.
How to start your first 90 day sprint
Pick the next quarter that is already coming. Not "the quarter after we have had time to plan properly." The next one.
Write three outcomes as measurable statements. Not "improve our SEO" but something like: "Move average position on twelve priority keywords from 19 to under 12 by end of quarter." Not "create more content" but: "Publish six new pieces in the priority topic cluster with rank tracking live on all target keywords." The measurement is the commitment. Vague statements produce vague sprints.
For each goal, name the one leading indicator you will check weekly. The indicator tells you mid-sprint whether you are on track, not retrospectively whether you succeeded.
Book four meetings before the sprint starts: kickoff, monthly check-in one, monthly check-in two, end-of-sprint review. In the calendar now. The meetings are not optional extras added when convenient. They are the sprint.
If you are not sure what realistic goals look like for your site and your team, the next section is for you.
Where Creative Tweed comes in
Every Creative Tweed engagement runs on a 90 day sprint cadence. It is not a methodology we borrowed from a book. It is the operating rhythm that makes the search visibility and traffic programme something you can measure and hold to account quarter by quarter rather than admire in a slide deck.
The Traffic Projection Report is a practical starting point if you want to see what a first sprint could realistically deliver for your site. It maps your current organic traffic picture, identifies which keyword positions are worth targeting in a first 90 days, and gives a projection for what movement in those positions would mean in actual sessions and leads. It is the analysis we run before starting work with any new client.
Free resource: Traffic Projection Report
See what a first 90 day sprint could deliver for your site.
90 day sprints are not magic. They are a cadence. The point of the cadence is that it forces honest measurement. Honest measurement is what most marketing programmes are missing.
Frequently asked questions
What is a 90 day sprint in marketing?
A 90 day sprint is a quarterly marketing operating rhythm where a small number of outcomes, usually three, are committed for the next quarter. Monthly checkpoints test progress, weekly stand-ups maintain pace, and the quarter ends with a structured review and the start of the next sprint. It operates at a timescale short enough to respond to real market change, rather than tracking against a plan written before the year started.
Why 90 days and not 30 or 12 months?
Thirty-day cycles are too short for SEO and content work to show measurable movement in the data. Twelve-month plans are too long to stay accurate in a marketing environment where search behaviour, AI platform behaviour, and channel costs are all shifting quarter to quarter. Ninety days is long enough to ship and measure something meaningful, and short enough to course-correct without losing a year.
How do I run a 90 day sprint for my marketing programme?
Pick three outcomes, write them as measurable statements, identify the one leading indicator for each goal, and book four meetings before the sprint starts: a kickoff, two monthly checkpoints, and an end-of-sprint review. Start with the next quarter that is already coming.
What should I measure in a 90 day marketing sprint?
Focus on leading indicators: indexed page count, average keyword position on your tracked set, AI citation appearance rate, organic landing page sessions, and lead magnet downloads. Revenue and conversion metrics matter but move on a longer arc than one quarter. Track them without expecting the sprint alone to shift them.
