We Rebuilt Flaunt For The Way People Actually Search

For over three years we have quietly rebuilt our proposition, our operating model and our tooling around a shift the rest of the industry is only now openly discussing. This is the full story of what we built, why we built it, and what it means for the brands we work with.

The conversation has finally caught up with us.

Across LinkedIn, conference stages and agency pitch decks, the industry has spent the last six to twelve months discovering that search has changed. That “discovery” is no longer a single box on a single platform. That the old playbook of ranking a page on Google and waiting for the click is quietly breaking. Senior marketers are asking the same question out loud: what is the playbook for digital growth now?

It is a good question. It is also one we started answering in 2020, and one we have been building toward, mostly in silence, ever since.

We have a habit at Flaunt of building first and talking later. It comes from being in rooms where unfounded promises were made. I would rather ship something that works than announce something that doesn’t, it also comes from a discipline that values evidence over noise. That instinct has served us well. But it has a cost. We have built one of the most complete responses to the discovery shift of any agency our size, and almost nobody outside our client base knows the half of it.

So this is me deciding now is the time to be louder. Not because the work is finished, but because the moment has arrived and the brands trying to navigate it deserve to know there is an agency that saw it coming.

This is Orbit, and the intelligence suite underneath it.

For twenty years, search optimisation meant one thing. Get a page to rank on Google, earn the click, convert the visit. The journey was tidy and the tooling was built around it.

That journey no longer describes how people behave. Your customers now look for answers, products and services across a sprawl of platforms. They ask ChatGPT. They watch a review on YouTube. They scroll TikTok and Instagram. They check Reddit threads. They read AI Overviews at the top of a Google result and never click anything at all. Google is still a central pillar, but it is one pillar among many, and demand and intent vary enormously across them.

One of the most useful ways we have found to describe this came from joint research by Boston Consulting Group and Google, which replaced the old awareness to loyalty funnel with something that matches reality. People move between four behaviours, in no fixed order: stream (where they watch), scroll (where they browse social), search (where they look things up) and shop (where they buy). One person passes through all four in a single afternoon. A funnel that assumes a neat sequence describes none of it.

Layer onto that the rise of AI mediated discovery and the shift becomes structural rather than cosmetic. The large platforms, Google, ChatGPT, Meta, etc, are competing to keep users inside their own ecosystems, because they monetise that attention better than they monetise passing traffic on. The answer increasingly arrives paraphrased inside a model, and the click, if it comes at all, comes later and further down the funnel.

Here is the part most people get wrong. This does not kill websites and brand experiences. It reframes them. Traffic that arrives via an AI citation lands with higher intent, which makes on site experience and conversion optimisation more important, not less. Commercial and transactional intent behaves very differently from informational intent, where zero click has been eating traffic for years. People still need to compare, buy, book and enquire. When that traffic clicks through, it is higher quality than it has ever been.

The work does not disappear. The shape of it changes. SEO becomes about being present inside the AI answer rather than ranking above it. CRO becomes more central because the visitor is closer to deciding. Brand becomes the layer that survives a platform paraphrasing you, and the thing most likely to get you cited inside an answer in the first place.

That is not an add on insight. That is the reframe. And reframing an entire agency around it is not something you do with a press release.

Most “AI first agency” announcements over the last two years are the same thing, wearing a new badge. Existing teams, existing service lines, existing deliverables, now with a thin layer of model calls wrapped around the same shape of work. There is nothing wrong with adopting AI tools. But bolting AI onto a pre-AI proposition is not the same as redesigning what the agency does around how behaviour has changed.

Retrofit agencies talk about AI as a capability they have added. The frame implies you could switch the layer off and the agency underneath would be unchanged. We talk about AI as the lens we rebuilt through, which implies you cannot switch it off, because the underlying disciplines have already been reshaped.

The reason that distinction matters is timing, and timing is the one thing you cannot retrofit. We started building tooling on AI and natural language processing in 2020, before most of the industry had a view on it. By the time competitors were commissioning their first AI generated content, we had already begun rewriting what the agency does, why it does it, and the instruments it runs on.

Building proprietary tooling and a genuinely connected proposition is a long game. You have to build the instruments, embed them in how teams actually work, and let them mature across real campaigns. It’s not something you can knee jerk, or repackage; it takes time.

Orbit is our answer to all of this. It is our framework for user centred search, and the simplest way to describe it is this: we put your audience at the centre, then build connected strategies that position your brand wherever those people are actually looking, across every searchable platform.

That sounds straightforward. Making it real required us to dismantle the thing most agencies are quietly built on: silos.

In a lot of traditional agencies, SEO, paid media, digital PR, organic social, creative and content sit in separate teams, working to separate plans, reporting on separate numbers. The client ends up as the integration layer, stitching six strategies into one and absorbing the overlap and wasted spend that comes from teams that do not share data. Orbit removes that. Our specialists across SEO and GEO, content, digital PR, paid media, organic social, creative, and web development work within a single strategy – sharing insight and data in real time. Every channel is connected, which means no duplication, no wasted spend, and growth you can actually measure.

A few principles hold it together:

It is platform agnostic. We are not wedded to Google (although Google is still hugely important), or to any single channel. We deliver the right mix to achieve the maximum commercial impact for a given brand, whether that is AI search, traditional search, social, performance or on site work, and we are honest when a platform is not where your audience is.

It is built on a strategic view of where influence actually happens, not where it is easiest to report on. We analyse performance opportunities and touchpoint influence across platforms to identify where each keyword, topic and channel delivers the greatest commercial impact, which is a very different exercise from optimising every channel in isolation and hoping the sum adds up.

And it is built on evidence, not opinion. Orbit runs on a constant cycle of testing, performance monitoring and forecasting. We evolve strategies based on what the data shows, keeping clients ahead of competitors and ahead of changes in search behaviour rather than reacting to them after the fact.

Orbit is already in market. It is not a concept we are floating. It is how we plan, execute and measure.

Orbit did not arrive fully formed. It is the front edge of an evolution that has run through the whole life of the company, and the rebuild of the last three years only worked because of the decade of discipline underneath it.

We were founded in 2015 with the simple goal of delivering better specialist service than we had seen in the market, and we built a reputation for technical quality and craft before we built anything else. We always self-funded our growth and built properly, this is no different, but it feels more significant. 

Through the pandemic, while many agencies contracted, our mixed delivery model grew. We expanded the service offering and built out the talent to support it. Then, from around 2021, we began developing proprietary frameworks and IP and started embedding AI into our core processes rather than treating it as a side experiment. We invested in the team, because a proposition this connected needs people who can hold the whole picture.

The last stretch, from 2023, is when the rebuild became visible internally. We launched Orbit. We rolled out company wide experimentation, including a public failure rate that we publish openly, because an agency that tests honestly should be willing to show how often its tests fail. We built an internal AI tooling that every team can use. We standardised delivery operations and raised the bar on the quality and consistency of what we ship.

None of that is glamorous. All of it is the operating discipline that AI tooling has to plug into in order to mean anything. The shared vocabulary, the standard reporting, the senior confidence in what is being measured, the data stack that gets richer with every campaign. Those do not ship with a model subscription. They take years, and we spent them.

The clearest evidence that we built rather than retrofitted is the timeline of our own tooling.

In 2020, we were experimenting with the OpenAI beta, working out what these models could do if applied properly to our work and our sector. In 2021, we built Goose, which remains our most mature piece of IP, created to enhance content production at scale. Through 2022 and 2023, we integrated Google’s natural language processing engine into Goose so it could pre optimise content the way a search engine interprets it, and we built our first forecasting tool on Meta’s Prophet machine learning model.

From 2023, we pushed AI adoption across the whole business, not just the technical teams. We ran internal initiatives to build AI fluency among staff, and we built FlauntGPT, an internal tool that makes prompt engineering and our best AI workflows accessible to everyone in the company. AI is now embedded in most of what we do, and we keep pushing it for one reason: it makes delivery more efficient and more consistent, which frees our people for the craft, creativity and strategy that actually moves the needle for clients. 

That is six years of compounding work. It is not an endpoint. It is the surface the next chapter operates from.

If Orbit is the proposition, influence mapping is the compass that provides informed directions for a specific brand.

An influence map and factor analysis give you insight into market perception across key channels, but also surface where to push. That is worth knowing, and it’s something we can measure consistently. This principle gives us the ability to shape budget distribution outside short term thinking and away from the messy middle of digital channel attribution. 

To do this we take the BCG and Google behaviour map, stream, scroll, search and shop, and across each behaviour we score a brand on four factors we can actually pull as levers on its behalf: reach (are you present where the audience is), attention (do you hold them once you are), relevance (are you the right answer to what they are doing) and trust (do they believe you). Four behaviours, four factors.

The value is not the headline number. It is in the small figures. Two brands can post the same overall influence and need completely opposite things, because one is weak on reach and strong on trust and the other is the reverse. Reach is a media problem. Trust is a brand and PR problem. Different owners, different budgets, different timelines. A single blended score hides exactly the thing you wanted to know. The map surfaces it.

The output is a clear, computed benchmark built from multiple data inputs, including Ahrefs, SimilarWeb,a mix of first-party data, and owned and paid social data. We named this PathwaveIQ, one of the tools in our intelligence suite, and we use it as a strategic jumping-off point and a quarterly benchmark. It does not replace the strategy. It tells us where to point it.

This is the part we have kept quietest about, and the part I now think we have undersold most.

Over the years, we have built a suite of proprietary tools. Individually, each one is useful. The reason they matter is that they are not individual. They connect, and together they form something most agencies our size simply do not have: a compounding data asset that gets richer with every campaign we run for a client.

This is what sits in the suite.

Goose is our content optimisation engine.

Goose was built in 2021 and now integrated with multiple AI models, including Google’s NLP engine, ChatGPT, Gemini and Claude. Our content team uses it to analyse entities and classification and to suggest concrete ways to improve a piece. Our digital PR team uses it to score the relevance of backlinks and coverage. It has run just under 47,000 optimisations to date. It tells us how relevant a piece of content actually is before it even goes live.

Bound is our digital PR intelligence tool.

Bound tracks coverage, sentiment and links, and combines SimilarWeb and Ahrefs data with our own algorithm to calculate an Estimated Coverage Value, a way of expressing the genuine commercial worth of PR coverage to the stakeholders who control budgets. Crucially, Bound measures off site content the same way Goose measures on site content, which closes the loop: we can score relevance everywhere your brand appears, not just on your own pages.

LLM Lens is our AI visibility tool.

LLM Lens allows us to generate AI native prompts that we can track in line with the topic cluster we optimise to. It also carries market intelligence across tracked prompts, which can be run against over 170 leading models. Where it gets interesting is our intelligent synthetic personas; we use Similarweb and Claude to create synthetic customers. This builds a detailed view of audiences that mirrors real life. We can then simulate a user’s journey with an LLM to analyse exact use cases. To bridge this all into commercial impact, LLM Lens also builds a knowledge map of a client’s site, highlighting content and topical authority gaps, structural flaws, and building a suggested PR campaign map to support offsite authority growth. The final piece of data is the multi-modal content audit, which uses Google’s Embeddings 2 model to analyse a site in parallel with how Google will view the multimodality of a site. 

Audience Lens is the inside-out counterpart.

Audience Lens builds personas and runs conversational analysis on your own Google Analytics profile data, so that after a campaign, we understand who your actual audience is and how they behave, rather than who we assumed they would be.

PathwaveIQ is the influence mapping tool described above.

Pathwave IQ draws on SimilarWeb, Ahrefs and SEOmonitor data to benchmark a brand across the behaviour map and the four factors, and it is the synthesis layer that turns the other tools’ signals into a strategic starting point.

Glass is the connective tissue of delivery.

Glass handles our dev reporting, client dashboards and content deployment, and gives clients genuine transparency into the work in progress rather than a monthly PDF after the fact.

Read that list again and notice what it adds up to. LLM Lens shows how the AI world sees you. Audience Lens shows who your audience really is. Goose and Bound measure relevance on site and off site in the same frame. PathwaveIQ benchmarks influence across the behaviours that drive your category. Glass keeps the delivery honest and visible. These are not six subscriptions we happen to own.
They are one connected instrument for understanding and growing a brand across the entire discovery ecosystem.

The honest part is this. We do not own the raw data. Almost nobody does; it lives inside the platforms, and the platforms are walling it off more every month. Our advantage is not access. It is what we do with it: how we analyse it, how we stack the signals together into an asset for a specific client, and how we use that stack to steer strategy through service delivery, which we have years of experience running. The more we work with a brand, the richer that asset becomes, and the more that data compounds into a guiding asset. 

There is a strategic choice buried in all of this that is worth making explicit, because many larger agencies have made the opposite one.

The temptation, when you have an R&D engineering budget, is to build your own closed AI platform and control everything inside it. The logic sounds right: build it ourselves, control it, customise it, stay ahead. The problem is that the AI bohemoths improve every single day, and no in house team keeps pace with that. Worse, a closed system locks you out of the ecosystem built around the models, connectors, plugins and integrations that actually make AI useful inside a real workflow. The agencies spending the most to look forward thinking are quietly engineering themselves into the next generation of legacy tech.

We made the other bet deliberately. We build our IP to plug into the ecosystem rather than wall it off. Goose runs across Google’s NLP engine, ChatGPT, Gemini and Claude rather than a single frozen model. LLM Lens reads across every engine that matters. When the frontier moves, our tools move with it, because they were built to ride the platforms rather than replace them.

We learned the edge of this the hard way, which is the most honest evidence I can offer. We once started building our own internal copilot that used retrieval to pull client insights from our files. We killed it before it shipped, because the off the shelf tools added the exact same capability while we were still building. The lesson was sharp: trying to innovate in front of the frontier models is close to impossible, given the speed they move. The tools worth building are the ones that sit alongside the leaders and connect their capability to a real client problem, not the ones trying to outrun them. Goose endures because it solves a specific content problem the models did not, and integrates with the rest of our stack. The copilot did not, so we let it go.

That discipline, building into the ecosystem and being willing to kill what the frontier overtakes, is why our suite remains relevant and continues extending rather than ageing into legacy.

Strip away the tooling names and the strategy language, and here is what Orbit and the intelligence suite actually deliver.

Right now, you get a single connected strategy instead of six disconnected ones, which removes the overlap and wasted spend that comes from siloed teams. You get a brand that is present where demand actually lives, across search, AI, social and commerce, rather than optimised for one platform while your audience drifts to others. You get conversion thinking applied at the exact moment it matters most, because the traffic arriving through AI discovery is higher intent and closer to deciding. And you get decisions made on evidence that quantifies the impact of every action on traffic, engagement and revenue.

Into the future, you get something rarer: a data asset that compounds. Every campaign we run makes our understanding of your brand and your category sharper, which makes the next decision better than the last. You get a partner that is platform agnostic by design, which means that when discovery shifts again, and it will, you are not exposed to a single channel or a single platform’s algorithm. And you get an agency that has already proven it can rebuild ahead of a structural shift, rather than one scrambling to respond after the fact.

As an illustration of where this leads, one of our fashion clients has seen its ChatGPT referred channel grow by triple digits year on year, with sessions, purchases and revenue all climbing fast. The absolute volumes are still small, because this is early. But no other channel is growing at that rate, and it is growing because we were measuring and optimising for it in the right way ahead of the curve.

We have spent the better part of a decade arriving at this point, and three intense years rebuilding the proposition, the operating model and the tooling around the way people actually search now. We did it quietly, because that is how we are wired, and because we would rather prove a thing than announce it.

The market has now caught up to the question we have been answering since 2020. Discovery has fractured across platforms, the old playbook is breaking, and brands need a partner who understands the whole ecosystem rather than a single channel. That partner exists. It is built, not bolted on. It is connected, not siloed. And it has the data asset, the tooling and the operating discipline to back every word of that up.

We were quietly ahead of the curve. We are done being quiet about it.

If you want to see what user centred search looks like for your brand, that is exactly what Orbit and our intelligence suite are built to show you.