Build the 6-month roadmap
Sequence, assign and estimate the actions over six months.
A ranked list tells you what matters. The roadmap says what happens in which month, who does it, what it costs and how you will know it worked. Most audits die between those two documents, because forty actions get handed to a team that can absorb a fraction of them.
I build it over six months. Any shorter and content, SEO and retention work can’t show a result before the deadline. Any longer and the assumptions you write down today will be stale by the time you reach them.
Throughout this chapter, say a B2B SaaS at 2M ARR. Its audit found unreliable attribution, blog CTAs nobody sees, no real bottom-of-funnel pages and a weak onboarding. All the figures below are illustrative. The starting point is the ranked list from the prioritization chapter.
Measure your capacity before you place anything
Count the days each person can really give to the roadmap, after what already runs: support, recurring campaigns, bugs, the newsletter. Then hold back 20% for what will come up anyway.
| Role | Days per month | Over six months | Plannable at 80% |
|---|---|---|---|
| Head of growth | 10 | 60 | 48 |
| Content lead | 12 | 72 | 58 |
| Developer | 10 | 60 | 48 |
| Freelance designer | 3 | 18 | 14 |
| Total | 35 | 210 | 168 |
The total is the least useful line. A plan can fit in 168 days and still fail because every action needs the same developer in the same month, so check the load per role and per month.
Sequence by dependency, then by score
Dependencies decide the phases. Scores decide the order inside each phase.
| Phase | Months | Purpose | Typical actions |
|---|---|---|---|
| Repair | 1 | Make the numbers trustworthy and stop the obvious losses | Tracking, UTMs, technical errors, invisible CTAs |
| Fix the bottleneck | 2 and 3 | Work on the stage the audit named | Bottom-of-funnel pages, CTA routing, links to feature and pricing pages |
| Compound | 4 to 6 | Build what keeps paying once it ships | Free tools, distribution workflows, success stories |
Tracking goes first even when a landing page scores higher, because every later action will be judged with it. Split it in two: the quick fix in month 1 (track the blog’s CTAs and navigation, add an acquisition analytics tool), the multi-touch model once the data is clean. The tracking chapter details both.
Write the dependencies down, since they are what people forget when someone asks to move a line.
| Action | Needs first | Because |
|---|---|---|
| Links from articles to feature pages | Feature pages worth landing on | Otherwise you route readers to a dead end |
| Multi-touch attribution | UTMs on CTAs, menus and alerts | The model can only weigh the touches it sees |
| Judging the new landing pages | Qualified leads reported by source | Signups alone would flatter the wrong pages |
Then place the fixed dates: a pricing change, a trade show, a seasonal peak. Those actions are planned backward from their deadline. On Grosbill, the dedicated Black Friday pages went live several months ahead, and SEO revenue over the period was multiplied by 4. Published in November, the same pages would have had no time to rank.
The grid that becomes the working document
One line per action, with one owner per line. Days are split as head of growth / content / developer / designer.
| Months | Action | Owner | Days | Success measure |
|---|---|---|---|---|
| 1 | Track blog CTAs, menus and alerts, set the UTM convention | Head of growth | 3 / 0 / 3 / 0 | Share of signups with a known source |
| 1 | Add acquisition analytics, report qualified leads by source | Head of growth | 4 / 0 / 2 / 0 | Report delivered and trusted by sales |
| 1 and 2 | Fix 404s, canonicals, hubs left in noindex | Developer | 0 / 0 / 8 / 0 | Errors at zero on the next crawl |
| 1 | Move CTAs up the page, send MOFU to BOFU | Content lead | 0 / 8 / 0 / 2 | Click rate on tracked CTAs |
| 2 to 4 | Build alternatives and versus landing pages | Content lead | 0 / 20 / 6 / 6 | Qualified leads from these pages |
| 2 and 3 | Link articles to feature and pricing pages by topic | Content lead | 0 / 6 / 0 / 0 | Visits to those pages from the blog |
| 3 and 4 | Rework onboarding toward the moment of value | Head of growth | 12 / 0 / 10 / 0 | Activation rate by signup cohort |
| 4 | Multi-touch attribution, 40% first, 40% last, 20% shared | Head of growth | 6 / 0 / 4 / 0 | Model used in the monthly review |
| 4 to 6 | Free tool with email capture | Developer | 5 / 0 / 12 / 4 | Emails captured, then qualified |
| 4 to 6 | Distribution workflow per content type | Content lead | 6 / 6 / 0 / 0 | Workflow applied to every publication |
| 1 to 6 | One success story a month | Content lead | 0 / 12 / 0 / 0 | Six stories, segmented by country |
| 1 to 6 | Monthly review | Head of growth | 6 / 0 / 0 / 0 | Decisions logged each month |
Sum the columns: 42 days for the head of growth, 52 for content, 45 for the developer, 12 for the designer. That makes 151 of the 168 plannable days, with every role under its own limit. Whatever doesn’t fit moves to the parked list below, however good its score.
Put a number on each line
Costing has two sides. For the cost, multiply days by a loaded day rate and add tools and freelancers. At a €500 loaded rate, this plan costs €75,500 in time. The landing pages alone, at 32 days, cost €16,000.
For the return, estimate a range and write down the assumption behind each figure. Close rates and contract values come from your own CRM, never from an industry average.
| Landing pages | Low | Expected | High |
|---|---|---|---|
| Added qualified leads per month | 3 | 8 | 12 |
| Close rate | 10% | 20% | 25% |
| Annual contract value | €12,000 | €12,000 | €12,000 |
| New ARR added per month | €3,600 | €19,200 | €36,000 |
The low case is the one to read. Here it adds €18,000 of ARR in five months against €16,000 of cost, which is a defensible bet. If your low case can’t justify the line, keep it only if you are ready to call it a bet in front of the team.
Add a stopping rule at the same time, while nobody is attached to the result: if the pages bring fewer than three qualified leads a month by the end of month 6, stop producing new ones and rework those that exist.
Write down what stays off the roadmap
On an audit I ran for a B2B SaaS, I stated what I had not dug into and gave no recommendations where I wasn’t the best placed. The roadmap needs the same honesty, so give the parked items their own table.
| Parked | Why | Reopen when |
|---|---|---|
| Second language | No proof yet that the current language converts | Bottom-of-funnel pages have proven themselves |
| Generic paid search | Attribution can’t judge it yet | Multi-touch has run for two months |
| Pricing review | Outside this audit’s mandate | The founder commissions it |
This table protects the plan when a new idea shows up in month 2. The idea goes on the list with a reopening condition, and the month keeps its shape.
Schedule the checkpoint now
Block a half day at the end of month 3. Replace estimated scores with measured results, cut what isn’t delivering, and pull forward anything from the parked list whose condition is now met. Months 4 to 6 stay a draft until this half day turns them into a plan.
What you deliver is one spreadsheet with four tabs: capacity, grid, costing and parked items. Your next step: fill the capacity table with real days per person, then place your ranked actions phase by phase until a role hits its limit. What remains goes to the parked tab, and the result is what you take to the presentation.