Dolphin Analytics turns disconnected marketing data into decisions. We reconcile GA4, Shopify, the ad platforms and your CRM into one set of numbers, then answer the questions that matter: which platform to trust, which channel is the growth lever, what to do next.
Axon Garside 
connective3 
AKQA 
WPP 
Omnicom 
Toyota 
Volvo 
Cedar 
Preqin 
Mediahub 
Tesco 
Heidi 
ElevateJet 
Illumicrate 
Octopus EV 
Noisy Beast 
Incisive Media 
Boltahead 
Robot Zebra 
Pixated 
Passionfruit 
Ada Ventures The same data problems come up for almost every marketing team we meet, agency-side or in-house. They usually sound like this.
The data exists, somewhere, across GA4, the ad platforms, the CRM and the store. Linking it into one answer is the part nobody has time for.
The board wants to know what marketing returned. The dashboards claim plenty, the bank account says less, and there's no defensible number in between.
Analysts, data scientists and data engineers are slow to hire and expensive to keep, and one hire rarely covers all three skill sets.
If any of these sound like your week, start the free audit. Tell us what's going wrong, by voice or text, and we'll come back with what's actually broken and what it would take to fix.
The concrete work behind the pillar, in plain terms.
We compare what your ad platforms claim against what your store and CRM actually recorded, and give you the one number the business can run on.
Which channels genuinely drive revenue, measured against real orders rather than dashboard claims, so budget decisions survive questioning.
The role each channel plays and where extra spend actually moves revenue. The growth lever, found and evidenced.
Modelling, forecasting, CRO and A/B testing when the question needs more than a report. Proven before the budget moves.
None of them alone. Each ad platform grades its own homework, so Meta, Google Ads and GA4 can all claim the same conversion. A source-of-truth reconciliation compares what the platforms claim against what the store actually recorded, and decides which number the business runs on.
The gap is bigger than most teams expect. For one subscription box brand, an ad platform claimed 6.9 times the revenue Shopify actually recorded, and 71% of the conversions it claimed came from people who saw an ad but never clicked it. The same reconciliation showed the channel was still genuinely profitable: a 4.7x return confirmed against orders, versus the 32.8x on the dashboard. Truth here wasn't doom; it was a number you could plan on.
Channels play roles. Some prospect, some capture demand, some retain, and judging them all on last-click ROAS punishes the prospecting that fills the funnel. Dolphin's Channel Insights work identifies the role each channel plays and which one is the growth lever, reading MER alongside ROAS.
From one brand's Channel Insights deck: email converting at 8.9%, roughly 3 to 4 times any other channel; subscriptions making up 32% of UK and 71% of US revenue while staying invisible to every marketing tool; and a book-launch campaign confirmed at a 9.4x return in 26 days, measured against Shopify orders rather than dashboard claims.
The data exists; it just lives in disconnected platforms with no shared definitions. GA4 holds sessions, Shopify holds orders, the CRM holds deals, and nothing agrees on what a customer is. Linking them is the actual work, and it's the part in-house teams rarely have time for.
One private aviation client had answers sitting in their CRM all along. Of 500 contacts with deals, 88.6% had no web attribution at all, and among the contacts that did, organic search drove 50.9%. The same linking work traced a $767,500 charter enquiry back to a single organic search visit. None of that was visible until the platforms were joined up.
Draft narrative
"Meta revenue up 23% month on month"
A human analyst reviews every deliverable
Yes, when every number is machine-checked against the source data before it ships. That validation step is the difference between analysis and hallucination: our validator once caught a report claiming 23% where the source data said 21.4%, and blocked it. A human analyst stays in the loop on every deliverable.
Experimentation is the supporting pillar under Insight: test and learn, CRO and A/B testing. Insight tells you what's likely to work; an experiment proves it before the budget moves. We design the test, size it honestly and read the result against revenue, so the learning compounds instead of sitting in a deck.
Agencies and in-house teams we plug into as their analytics bench, on what it's like to have us on the team.
Partnering with Dolphin Analytics resonates with us because it gives us extra capacity with heavily fluctuating demand in this current climate around new tracking implementation technologies. It's the offering of dealing with experts with a no-fuss approach. They're so valuable, solving our biggest pain points by providing on-demand expertise. In any agency setting there is always a lack of capacity for niche skilled expertise, and Dolphin Analytics provides that.
Alan Ng Chief Innovation Officer, Connective3 Marketing agency, 150 people Working with Dolphin Analytics has been a pleasure. Olam and his team were brought on to help us work through client issues outside our usual area of operations. Their team worked seamlessly alongside Axon Garside for our clients. They led the project, identifying the solution swiftly, before developing the fix and completing the jobs rapidly. Dolphin Analytics' communication was great throughout, keeping my team updated. This is the beginning of a fruitful business relationship between both businesses.
Spencer Montagu Marketing Director, Axon Garside B2B marketing agency, 20 people Olam and his team joined us when our product analytics was still in its infancy. Within weeks they'd resolved a host of issues with our Heap events and reports, giving us a much clearer picture of how customers used the platform. They then helped our product and UX team embed a framework for experimentation, so we could validate ideas and deliver value far more quickly: a real step change in how we build products. Olam is a genuine people person, and his collaborative approach was hugely appreciated. I extended the engagement numerous times and am hugely grateful for the change he helped enact.
Jack Levy Head of Product, Preqin In-house product team Data can be complex and, more often, it is made unnecessarily so by the people within it. However, Dolphin Analytics' skill is in understanding what's possible, what's necessary, and making that easily understood and implemented. It's rare to have such consistently good and reliable data expertise, but Dolphin Analytics always delivers.
Dino Myers-Lamptey Managing Director, Mediahub Marketing agency, 500 people Dolphin Analytics joined our team and quickly became a valuable asset that we wanted to extend onto other projects. They are able to think critically, stay solutions-oriented, and know how to work with others, bringing their expertise and point of view on data and analytics. Beyond the technical skill across various analytics tools, they have broad experience with different types of analysis, which makes them able to step up to different challenges.
Payam Cherchian Data & Analytics Director, AKQA (now EssenceMediacom) Web & design agency, 40 people We turned to Dolphin Analytics when we had a data project we needed specialist input on. The team took time to understand our challenges, ambition and the nuances of our business to propose a bespoke solution. They broke down complex processes in a way we could communicate across our teams and clients to build into our roadmap.
Joseph Costello Global Strategy & Insights Director, Cedar Swipe / scroll → 1 / 6
Dolphin Analytics has helped an agency avoid exposure to a potential GDPR fine of up to €20M, saved a media agency 20 hours a week on reporting, and shown brands which of their ad numbers reflect reality.
Connective3 · Performance agency
20+ projects
delivered · 5 clients
Tracking, visualisation and insight delivered across their client roster, helping keep retainers worth multiple six figures combined.
How: embedded team · tracking fixes · Piano Analytics · dashboards
Toyota · via AKQA
3x conversion rate
after the repaired journey
People were dropping out through hidden gaps in the journey. The work found the weak points, closed them with experiments, and kept more of the same traffic moving to the end.
How: insight · conversion funnel repair · experimentation
ElevateJet · Private aviation
<3 months
from zero to 8 channels wired
Tracking, CRM attribution, BigQuery and Looker dashboards from a standing start; their agency's performance finally provable, and a new partner onboarded on the evidence.
How: 3 sites → HubSpot · BigQuery + Looker · new-agency onboarding
Illumicrate · with Noisy Beast
5x ROAS
Confirmed against Shopify sales
We implemented a new tracking infrastructure and analysed performance across channels. The work confirmed Meta was profitable, giving the team confidence to increase spend and use better measurement to grow the waitlist.
How: full tracking setup · cross-channel analysis · profitable Meta spend confirmed
Heidi · Ski travel
2 weeks
From zero to launch-ready dashboards
Looker Studio built from scratch for a team with no data resource, ready in two weeks to run Black Friday on live numbers and see what to promote.
How: Looker Studio from scratch · launch support · no internal resource needed
Analysis is only as good as the tracking underneath it. Sonar's free scan shows what's actually firing on your site, so we know whether the numbers we're reconciling can be trusted at the source.
Why it exists: you can't fix tracking you can't see. Sonar shows us both the truth before any work starts.
Scan your site
Surface is our analytics platform: the reporting pillar, productised. Reporting used to eat a week a month and still left three numbers that disagreed. Surface gives the week back and settles the argument.
Every Dolphin Analytics engagement starts with a free audit. Tell us what's going wrong with your data, by voice or text, and we come back with a considered response. From there it's a scoped project or an ongoing value-based retainer.
Type or talk us through what's going wrong, what you've tried and why it hasn't worked. Three open questions, no booking forms, no sales call to sit through.
We read the transcript, look at the problem properly and come back with a considered response. Sometimes a call helps us clarify the problem. Sometimes we need view-only access to the relevant systems to see what's happening for ourselves. We'll tell you what we need and why.
Pricing is value-based on both: you pay for the outcome, not the hours.
We fix one scoped thing.
We run your analytics, ongoing.
None of them alone: each platform grades its own homework, so Meta, Google Ads and GA4 can all claim the same sale. A source-of-truth reconciliation compares platform claims against actual store or CRM records and decides which number the business runs on. For one brand, that cut a 32.8x dashboard claim down to a confirmed 4.7x return: still profitable, finally believable.
There is no universally right attribution model. Last-click undervalues prospecting, while data-driven models spread credit but hide their working. Pick the model that matches the decision you're making, and check the platform's claims against real orders before trusting any of them.
Channel-role analysis. Channels do different jobs (prospecting, capturing demand, retaining customers), so judging them all on ROAS alone punishes the ones that fill the funnel. Dolphin Analytics reads MER alongside ROAS and channel roles to find the growth lever: the channel where extra spend or effort actually moves revenue.
Yes, with validation in the loop. Every number in a Dolphin Analytics report is machine-checked against the source data before it ships, and a human analyst reviews every deliverable. That validation applies to our reports and analysis; it's the difference between answers and hallucinations.
Reporting tells you what happened; insight tells you what to do about it. The two hand off: reporting proves performance on schedule, insight finds the why and the next move. Dolphin Analytics treats them as separate pillars because mixing them is how reports get long and decisions go missing.
Yes. Experimentation sits under the Insight pillar at Dolphin Analytics: test and learn, CRO and A/B testing. Insight finds the likely win, an experiment proves or kills it, and the result feeds the next round.
GA4 audit checklist covering data streams, key events, Consent Mode, DebugView and BigQuery, in the order that catches the most tracking faults first.
12 Aug 2026Meta ads overreporting conversions: why Meta Ads Manager shows more sales than Shopify and GA4, how view-through inflates it, and the reconciliation fix.
11 Aug 2026How to choose a Google Analytics agency: the GA4, GTM and consent questions to ask, typical pricing, red flags, and when to hire versus keep it in-house.
Talk it through or type it out. Three questions, a few minutes. We read every transcript ourselves and reply with how we can solve your problem.
No calendar. No sales script. We review every response before suggesting a call.
Tell us which numbers you don't trust, what looks broken, and where you see it: GA4, Meta, Shopify, HubSpot, or somewhere else.
Include internal fixes, agencies, freelancers, or tools. What changed, what stayed broken, and why do you think it failed?
Is there a deadline, launch, reporting cycle, or campaign spend at risk? Tell us what fixing it would change.
Prefer email? olam@dolphinanalytics.co.uk.