One of the potential frustrations with any AI visibility monitoring tool is fairly obvious.
It can tell you that a competitor appears in AI-generated answers more often than you do. It can show you which websites are being cited, identify prompts where your business is missing and track whether your visibility is improving or declining.
But eventually a small business owner is likely to ask a much more practical question:
What do I actually do about it?
This is particularly interesting in the case of OtterlyAI because there is evidence that this was a genuine weakness users identified — and Otterly subsequently developed a feature specifically intended to address it.
It also connects with something I noticed during my recent test of OtterlyAI’s own visibility in Google AI Mode.
As I changed the emphasis of the conversation, the platforms AI Mode recommended changed too. In one response where I became particularly interested in understanding why a competitor might be outperforming a business and what could be done about it, Otterly disappeared from the recommendations.
One AI Mode response proves very little. These answers are dynamic, and running the same or similar query again can produce a different result.
But it raised an interesting question:
Is OtterlyAI mainly useful for diagnosing an AI visibility problem, or can it now help a business decide what to do next?
There is evidence that the answer may be changing.
The Problem: Visibility Data Doesn’t Automatically Tell You What to Do
Imagine that you use an AI visibility platform and discover that:
- your business appears in 20% of the AI responses you are monitoring;
- a major competitor appears in 60%;
- particular websites repeatedly get cited;
- the competitor appears for important prompts where you are absent;
- some of your pages receive citations while others never appear.
That’s potentially very useful information.
But it still leaves a gap between:
“Here is what is happening.”
and:
“Here is what I should investigate or change.”
Interestingly, Otterly itself acknowledged this problem when it launched Recommendations on 16 April 2026.
In OtterlyAI’s announcement of the Recommendations feature, CEO Thomas Peham explained that customers could already see brand coverage, citations and competitor visibility but were asking what they were actually supposed to do with the information.
Otterly described this as the gap between insight and action, and Recommendations was designed specifically to address it.
That is important.
This isn’t simply an outside reviewer deciding that Otterly needed more actionable guidance. Otterly itself recognised that collecting and displaying visibility data was only part of the job.
What Does OtterlyAI Recommendations Actually Do?
Recommendations sits inside an Otterly brand report and analyses the data Otterly has collected.
Instead of leaving the user to interpret all that information alone, Otterly attempts to identify opportunities and suggest actions.
According to Otterly’s current Recommendations documentation, each recommendation contains two particularly useful elements:
Reasoning — why Otterly has identified something as an opportunity based on the available data.
Suggestion — a concrete action that could be taken in response.
Recommendations are also prioritised by estimated impact as High, Medium or Low.
That is already quite different from a dashboard simply telling you that your competitor has a higher share of voice.
The platform is attempting to move from reporting what it found towards interpreting what the business might do about it.
The Competitor Gap Recommendations Are Particularly Interesting
Some of the current recommendation triggers are especially relevant to the question that started me thinking about this.
Otterly can identify situations such as:
- a competitor being cited on a blog or brand website when your business isn’t;
- a competitor being cited in Reddit conversations where your business isn’t;
- competitors appearing in media sources where you don’t;
- competitors receiving citations from YouTube;
- competitors appearing for a prompt while your business receives neither a mention nor a citation;
- an existing page ranking in Google but failing to generate AI mentions.
Depending on the situation, Otterly can then recommend investigating content partnerships, Reddit participation, media opportunities, YouTube, creating new on-page content or improving existing content.
For me, this is an important distinction.
If I discover that a competitor is repeatedly appearing for an important customer query while my business isn’t, I don’t particularly need another chart telling me that the competitor is winning.
My next question is:
Why might that be happening, and where should I investigate?
Recommendations appears to be Otterly’s attempt to help answer that second question.
There Is a Simple Workflow for Acting on Recommendations
Otterly also gives users a basic process for managing the suggestions.
Recommendations initially appear under Suggested.
If you think one is worth pursuing, you can move it into To-Do, where internal notes can be added.
Once the action has been completed, it can be moved to Archive.
If a recommendation doesn’t make sense for the business, it can simply be discarded.
I actually think that last option is important.
An AI-generated recommendation should not automatically be treated as an instruction.
For example, Otterly might discover that competitors are appearing in Reddit citations and suggest investigating Reddit.
That does not automatically mean every business should start posting on Reddit.
You still need to ask whether the recommendation makes sense for your customers, resources, expertise and overall marketing strategy.
Otterly itself says businesses do not need to act on every recommendation and should focus on the ones that are relevant and valuable.
That seems a sensible approach.
You Need Enough Data Before Recommendations Become Useful
There are also some practical conditions worth understanding.
According to Otterly’s documentation, the first recommendations require:
- at least 15 prompts;
- at least three competitors;
- at least three days of collected data, equivalent to three completed runs.
Once those requirements have been met, Recommendations refreshes every seven days.
The feature is available across Otterly’s plans, but the level of access differs.
The Lite plan provides a preview of up to three recommendations in each seven-day cycle, while Standard and Premium customers receive full access to the generated recommendations.
Otterly also currently describes Recommendations as being in beta and says that relevance and precision will continue to be improved.
That final point is worth remembering.
I would regard Recommendations as guidance about what deserves further investigation, rather than a list of changes that should automatically be made to a website.
But Did Recommendations Actually Fix the Problem?
This is where the user reviews become particularly interesting.
Recommendations launched on 16 April 2026.
Yet on 5 June 2026, almost two months after the launch, a verified current user called Anastasiia P. still highlighted the transition from insights to actionable recommendations as an area where Otterly could improve.
The reviewer liked Otterly’s interface, competitor identification and monitoring capabilities but felt it was not always clear what specific actions should follow from the findings, and wanted more concrete strategic guidance. G2 identifies the review as organic.
Readers can see that review, along with the newer reviews discussed below, on OtterlyAI’s G2 reviews page.
I think this June review is important because it stops us telling an artificially neat story.
It would be easy to write:
Users had a problem → Otterly launched Recommendations → problem solved.
The evidence doesn’t support such a simple conclusion.
Launching the feature did not mean every user immediately felt they were receiving enough guidance.
Perhaps the reviewer wasn’t using Recommendations extensively. Perhaps the early version didn’t meet their needs. Perhaps their expectations were different. Perhaps the feature subsequently improved.
We don’t know.
But we can look at what later users said.
The August Reviews Look Quite Different
On 12 August 2026, Dave M., another verified current user, described Otterly’s discoveries as making the next steps for GEO strategy and execution clear and actionable.
G2 identifies this review as organic.
Then on 19 August 2026, Sandeep J., also identified by G2 as an organic verified current user, specifically praised the newer Recommendations section and said it provided actionable pointers for improving GEO.
That gives us an interesting timeline:
16 April 2026: Otterly launches Recommendations specifically to address the gap between visibility data and action.
5 June 2026: a verified user still wants clearer actionable recommendations.
12 August 2026: a verified organic reviewer says Otterly now makes the next steps clear and actionable.
19 August 2026: another verified organic reviewer specifically praises the new Recommendations section for providing actionable pointers.
I don’t think those reviews prove that Otterly has completely solved the problem.
Three individual reviews cannot establish what every Otterly customer experiences.
But they do provide some evidence that the product is moving in the direction Otterly intended when Recommendations was introduced.
I Don’t Think It Is Accurate to Describe Otterly as Only a Diagnostic Tool Anymore
This is probably the biggest thing I have taken from looking more closely at the feature.
Describing Otterly primarily as a tool that tells you whether you appear in AI search, shows competitors and leaves you to work out everything else may once have been a reasonable simplification.
I don’t think it adequately describes the current product.
Otterly is still very much a monitoring and diagnostic platform.
But Recommendations adds another layer:
Measurement → diagnosis → suggested action.
That doesn’t mean Otterly actually implements those improvements for you.
There is an important difference between telling a business where an opportunity might exist and doing the work required to take advantage of it.
Recommendations Still Doesn’t Remove the Need for Judgement
Suppose Otterly identifies an important prompt where a competitor appears and your business doesn’t.
It might recommend creating new content or improving an existing page.
You still need to work out:
- what information customers actually need;
- what your existing page is missing;
- whether competitors provide information you don’t;
- what genuine expertise or evidence you can contribute;
- whether creating another page is really necessary;
- how any new content fits into the wider customer journey.
Likewise, identifying an opportunity for Reddit participation, media coverage, a partnership or YouTube content does not automatically mean pursuing that opportunity will be worthwhile.
The business still has to make the commercial decision.
Personally, I think that distinction is healthy.
I would be uncomfortable with an AI visibility tool automatically rewriting large parts of a website simply because its algorithm believed doing so might increase AI citations.
Showing me the evidence, suggesting what I might investigate and allowing me to decide what makes sense seems a more sensible approach.
This Brings Me Back to My Google AI Mode Experiment
The reason I became particularly interested in this subject was my recent experiment looking at whether an AI visibility platform was itself visible in AI Mode.
As I altered what the hypothetical customer needed, different tools became more or less prominent.
At one point the emphasis shifted towards understanding why a competitor might have beaten the business for an AI query and what could actually be done about it.
Otterly disappeared from that particular recommendation.
You can read the full experiment here:
Does an AI Visibility Tool Have Good AI Visibility? I Tested OtterlyAI in Google AI Mode
I want to be very careful about what we infer from that.
It does not prove that Google AI Mode sees Otterly as purely diagnostic.
It does not prove that reviews criticising a lack of actionable guidance caused Otterly to disappear.
And another run might have produced a different recommendation altogether.
But the combination raises a much broader question that I think is worth testing.
Can a Product Change Faster Than the Internet’s Understanding of It?
Businesses constantly change.
They introduce new services.
Software companies add features.
Pricing changes.
Different customer groups become important.
Weaknesses are addressed.
But the information describing those businesses across the web doesn’t necessarily change at the same speed.
Imagine a company that has historically been known for doing A and B.
Hundreds of webpages, reviews, comparisons and discussions describe it in those terms.
The company then introduces C.
Its own website is updated, but much of the wider information environment continues associating the business primarily with A and B.
Now imagine somebody asks an AI system:
“Which company can help me with C?”
Will that business be considered?
We cannot know exactly how any individual AI system makes that decision, and it would be wrong to assume that changing a few webpages would guarantee a recommendation.
But it does seem sensible for a business to make important capabilities clear, specific and well evidenced.
The evolution of Otterly’s Recommendations feature gives us an interesting real-world example.
The product itself has changed.
The question is how quickly the wider understanding of that product changes with it.
So, Has OtterlyAI Solved the “What Do I Do Next?” Problem?
I wouldn’t claim that it has completely solved it.
Recommendations remains in beta.
Otterly itself says relevance and precision will continue to improve.
And importantly, a verified reviewer was still asking for more actionable guidance nearly two months after the feature originally launched.
But there is also more recent evidence pointing in the other direction.
Otterly recognised that visibility data alone doesn’t necessarily tell a business what to do.
It built Recommendations specifically to tackle that problem.
The feature now uses collected visibility and competitor data to identify opportunities, explain why they have been flagged and suggest possible actions.
And recent verified users are specifically describing those discoveries and recommendations as actionable.
For a small business owner, I think that changes how Otterly should be evaluated.
It isn’t simply:
“Tell me whether my business appears in AI search.”
The proposition is increasingly closer to:
“Show me where I appear, show me where competitors are beating me, identify some of the gaps and opportunities behind those differences, and suggest what I could investigate next.”
Whether acting on those recommendations actually improves AI visibility is another question entirely.
And, as with most things in AI visibility, I think the best way to answer that question is through testing rather than assumption.