Core feature
Recommendations you can hand over, tasks you can prove you shipped
Findings only matter when someone changes a page. Recommendations turn evidence from scans, snapshots and comparisons into specific content work, and task tracking keeps an honest record of what was done, per client.
What it is for
The question it answers
“What should we change first?”
A list of two hundred issues does not tell a busy marketer what to do on Monday. Recommendations are written as practical actions: a section to add, a question to answer, a fact to state as text, a term your page never uses. Each one points back to the observation that prompted it.
Task tracking closes the loop. Mark a recommendation done when it ships, reopen it if it needs more work, and the completed list becomes a record you can share with leadership or a client. It also tells you which snapshots to re-run, because you know exactly what changed and when.
Inputs, workflow and outputs
How it works in practice
You bring
- Findings from website scans, ChatGPT snapshots and research tools
- Your judgement about which changes matter to real customers
- A writer, developer or yourself to publish the changes
Review recommendations
Read suggestions tied to a page, a comparison or an observation, with the reason it is an opportunity.
Copy and brief
Copy a full recommendation into a brief for a writer or developer, adapting it to your brand voice.
Mark it done
Once the change is published, mark it complete. Reopen it if the work needs another pass.
Review completed work
Use the completed tasks list to report progress and decide which observations to re-run.
You get
- Specific content actions with the evidence behind each
- A completed tasks history per client workspace
- Material for client and leadership reports
- A clear trigger for when to re-scan or re-run a snapshot
Who benefits
Who gets the most from it
- Content leads
- Brief writers with precise tasks instead of vague “optimise for AI” requests.
- Project managers
- See what is open, what is done and what was reopened, client by client.
- Agencies
- Show clients a concrete history of work completed, not just a monthly opinion.
Illustrative example
A software company with inconsistent pricing facts
Illustrative example, not customer data. A B2B scheduling app notices AI answers quote an old price.
A snapshot states the starter price as a figure the company retired last year.
Update the pricing page and two older blog posts so the current price appears consistently as text.
The writer updates the posts; the developer fixes the pricing table. Both tasks are marked done.
The original prompt is re-run later and the result noted. Change may take time and is not guaranteed.
Limitations
What it does not do
Honest limits help you set the right expectations with stakeholders. See the general FAQ for more.
- Recommendations are suggestions. Your team decides what is accurate, on-brand and worth publishing.
- The tool does not publish changes to your website or CMS.
- Completing tasks does not guarantee AI mentions or search rankings.
- Recommendations are only as good as the evidence behind them, so keep prompts and scans current.
Questions
Actionable recommendations and task tracking: frequently asked questions
Are recommendations generated automatically?
They are produced from your scans, observations and comparisons. They are starting points written for people to review, not copy to publish unchecked.
Can I reopen a completed task?
Yes. Completed and reopened actions are kept, so the history reflects what really happened.
Can clients see completed tasks?
Completed work can be summarised in reports. Access to the workspace itself depends on the team access you set.
Related features
No ranking guarantee. AI Search Console provides observations and recommendations only. It cannot guarantee rankings in any search engine or mentions in any AI-generated answer, which are determined by third-party providers. Product examples on this site are illustrative and do not show real customer data.