PlatformHow it worksPricingFull Intelligence ReportFor small businessEnterprise Contact
Sequencing

The fix roadmap:
the order is the plan

A list of problems is not a plan. OG01 orders findings by what each one is costing you rather than by how quickly it could be ticked off, then sequences them around dependencies — because some fixes cannot be measured until an earlier one lands. The result is a roadmap you can put against a quarter: the first item is genuinely the first thing worth doing, and the last item is honestly labelled as the thing that can wait.

Get findings to sequence → How the order is decided

The free reading gives you the top actions. The full reading gives you the whole ordered list with implementation detail.

Ordering

Impact first, not effort first

Most remediation lists are sorted by effort, because effort is easy to estimate and impact is not. The result is a quarter spent on twelve quick wins that move nothing, while the one structural problem underneath them stays exactly where it was. OG01 sorts the other way: what a gap is costing decides its position, and a hard item that matters outranks an easy item that does not.

That has a consequence worth stating up front. Your roadmap will sometimes open with something inconvenient — an entity inconsistency that spans systems you do not own, or a rendering decision baked into your platform. That is not the roadmap being unhelpful. It is the roadmap refusing to flatter you with an ordering you would have preferred.

Dependencies

What blocks what

Sequencing is not just ranking. Some items are prerequisites, and doing them out of order wastes the work rather than merely delaying it.

  • Retrievability comes before everything. If a page cannot be fetched and parsed, nothing measured on that page is trustworthy and no content fix on it can be verified. A Not Rated reading needs its explanation reviewed first. Use the Not Rated FAQ to decide whether to retry or ask for support before changing access settings.
  • Entity clarity comes before corroboration. There is no point earning external references to a business that resolves as three near-identical entities; the references land on the wrong one or split between them.
  • Structure comes before volume. Publishing more pages in a form nothing can quote adds retrieval surface without adding retrievable answers.
  • Off-site corrections start early and finish last. They depend on a third party acting, so they are begun in the first pass and closed out whenever the source responds — which is why they sit oddly in any plan sorted purely by impact.
Running it

A roadmap against a real quarter

Weeks 1–2 — clear the blockers
Anything preventing measurement: access, rendering, pages whose substance only exists after load. Nothing downstream is verifiable until these are done, so they are not negotiable regardless of appetite.
Weeks 2–6 — entity and structure
Make the organisation resolve to one thing and say what it is in a machine-readable form. Usually the largest single movement in a rescore, and the least visible work to anyone watching the site.
Weeks 3–10 — content that answers
Editing pages so the question a title asks is answered in a passage that stands alone. Frequently editing rather than writing, and frequently the part a marketing team can own without external help.
Ongoing — off-site corrections
Requests to directories, aggregators and third-party pages carrying stale or contradicted information. Opened early, closed as sources respond, never guaranteed because a third party has to act.

Then stop and let it recrawl. Rescoring the week the work ships mostly measures crawl lag; a few weeks after implementation is a fair first check for most sites.

Closing the loop

The rescore is what makes it a roadmap rather than a wish list

A plan nobody measures afterwards is just a list of opinions with dates attached. Because the second reading uses the identical method, the difference between the two is attributable to the work rather than to a changed measurement — and where nothing moved, the rescore says so. That is the uncomfortable half, and it is the half that makes the comfortable half worth anything.

Findings that did not move are not automatically failures either. An off-site correction that a directory has not yet applied, or a structural change that has not been recrawled, will read as unchanged for entirely mundane reasons. The roadmap distinguishes work that is done from work that is measurable, because conflating the two is how teams lose confidence in a programme that is actually on track.

What the findings look like → Having them executed → Why the two readings are comparable → The reading with the full roadmap → What it costs →
Questions

Roadmap questions

Can I reorder the roadmap to do the easy things first?
You can, and sometimes it is the right call for capacity reasons. Be aware of what you are trading: the ordering exists because impact and effort are unrelated, so an effort-first quarter can complete a lot of items and move the measurement very little. The dependency items are the ones not to reorder, because doing them late wastes the work that depended on them.
How many items are on a typical roadmap?
It varies with the domain, because the roadmap is generated from what the reading actually found rather than filled to a target length. A site with one structural problem may get a short list; a portfolio with inconsistent entity records across many locations will get a long one with a lot of repetition, which is exactly where a written-once specification pays off.
Does the roadmap tell me who should do each item?
It tells you what has to change and in enough detail to hand over, which usually makes the owner obvious — developer, SEO specialist, content editor, or whoever controls a third-party listing. It does not assume your team structure, because a specification that is written to be executed travels better than one written for a particular job title.
What if an item is impossible on my platform?
Some are, particularly rendering behaviour baked into a hosted platform. The honest position is that the finding stays on the roadmap as unresolved rather than being quietly dropped, so the rescore explains itself later. A finding you chose not to fix and a finding that was fixed and did not move are different results, and the record should show which is which.
When should I rescore against the roadmap?
Once the work has landed and been recrawled — a few weeks after implementation for most sites. Structured data and technical corrections can be picked up within days; anything depending on other sources corroborating you takes longer, and rescoring too early measures the lag rather than the work.
You need findings before you need an order for them.

The free score returns the highest-priority actions in about ninety seconds. The full reading returns the whole roadmap, sequenced, with the detail to hand each item over.

Get the free score → Full Intelligence Report — $399 →