Marketing Data Availability Audit · GIG Gulf

We cannot prove our marketing is working — because the data that would prove it isn't connected.

An honest audit of what marketing data GIG actually has, where it lives, and whether the team can reach it in the view they need, updated, without anyone doing it by hand. Built on GIG's own audit framework — available inside the source, available outside it, available in Power BI — with the blank cells filled in honestly. The headline: most of the data everyone assumes is "available" is in the building but out of reach.

This is not a pitch and it is not a request to buy a tool. It is a status check on a single question: can the marketing team show, with data, whether what we do makes a profitable difference? Today the honest answer is no — not because the work is weak, but because the numbers that would settle it sit in systems we can't see, can't reach, or can only stitch together by hand.

The trap is the word "available." Almost every figure anyone asks for does exist somewhere inside GIG. But existing in a system is not the same as being usable. This audit draws that line precisely, in GIG's own framework, source by source.

1 · The question we can't answer

When the natural question comes up in a meeting — "how much did we spend, and what did it return?" — marketing can answer the first half and only gesture at the second. We can show spend, clicks, leads, even GWP. We cannot reliably show whether that GWP was won profitably, by which channel, for which line of business, in which country. The data to answer that is not assembled anywhere a marketer can look at it.

That is the gap this document is about. It is a data-availability gap, and it is large.

The thesis in one line

Everyone believes the data is there. It mostly isn't — not where it counts, not in a form the team can use.

"In the company" is doing a lot of work in that sentence. Data can be inside GIG and still be invisible to the people who need it: locked in a system marketing can't open, gated behind a request, or arriving as an Excel that one person rebuilds by hand every week. None of that is available in any sense that lets the team do its job.

2 · What "available" really means

GIG's own audit deck already asks the right question, in three columns: is the data available inside the source platform, available outside it, and available in Power BI? That framing is exactly right — and this audit keeps it. The crucial point it surfaces is that only the third column counts. Data trapped "inside" Google Ads, or reachable only as a manual export "outside" it, is not available to the team in any usable way.

So a source is genuinely available only when it is democratically visible to the team, in the view they need, kept up to date, with zero manual intervention — i.e. a real, modelled "Yes" in the Power BI column. The states below describe what the first two columns actually mean on the ground.

State	What it actually means on the ground
In Power BI	Lives in a shared, modelled view, refreshes itself, and anyone on the team can open it and slice it. This is the only state that counts as "available." Today, almost nothing is here.
Outside · request-gated	The data exists, but to get it you have to ask a specific person or licence each time. Available in principle; not on your desk when you need it.
Outside · manual Excel	Exists, but only reaches the team because someone downloads it, builds the pivot, joins it by hand, and pastes it into a report. Available to one person's spreadsheet — not to the team, not automatically.
Inside only · no access	Trapped in a platform marketing can't export from, or in a system we can't open at all (policy admin, the call-centre platform). It might as well not exist for our purposes.
We don't know / not tracked	Either never captured, or captured without the keys that would let us join it to outcomes. Much of the competitor and offline picture sits here.

The single biggest dependency in the whole chain is one person rebuilding the weekly join by hand in Excel. That's not a connected data estate — it's a heroic workaround, and a single point of failure.

3 · Where we actually stand

Scoring ourselves honestly against that definition — counting only data that is genuinely in Power BI, automated and democratic — this is the picture. The percentages are Articulate's directional assessment, not a measured index; they're here to size the gap.

~25%

Today
where we are

40–50%

Ceiling of
current work

70–80%

What the job
actually needs

0% — flying blind100% — fully connected, profit-aware

~25% today. Paid tracking works inside the platforms; almost everything else is manual, gated or missing — and barely anything is in Power BI.

40–50% is the ceiling of the work currently in flight (the weekly reporting automation). It cleans up paid and organic reporting — but it cannot, on its own, link spend to profit or reach the data marketing can't access.

70–80% is what "doing the job" requires — profit-aware, all channels, no manual joins. The distance from 50% to 80% is not a marketing-tooling problem. It's a data-access and data-modelling problem, and most of it can only be unlocked inside GIG.

The important number is not 25%. It's the gap between 50% and 80% — the part the current workstream cannot reach no matter how well it runs, because it depends on systems and joins that sit outside marketing's hands.

4 · The audit, source by source

GIG's framework, completed. Each table reads across three columns — available inside the source, outside it, and in Power BI. Where the source deck left cells blank, we've filled them with the honest read.

Yes usable / connected Manual export by hand Partial some of it No not available ? we don't know

Paid — Google Ads / Bing / Facebook / X

Our strongest area. Note the shape: almost everything is available inside the ad platforms, and almost nothing makes it outside them or into Power BI. The data exists; it just never reaches a shared view.

Dimension	Inside source	Outside source	In Power BI
Campaign	Yes	Manual	No
Landing page	Yes	No	No
Sitelink	Yes	No	No
Device	Yes	No	No
Demographics — age, location, nationality	Yes	No	No
Day of week / hour of day	Yes	No	No
Ad type / banners / ad-copy performance	Yes	No	No
Ad group	Yes	No	No
Profit on Ad Spend (POAS)	Yes*	No	No
Forecast vs target vs spend	No	No	No
Optimise on sales	Partial	No	No

*POAS caveat. The platform can show a profit-on-ad-spend figure, but it's built on revenue/GWP proxies, not true policy profit — which needs eBao's car-type and value data that isn't joined in. So even where "Yes" appears, it can't yet mean profit. And it's neither outside the platform nor in Power BI, so no one beyond the ad account can use it.

Organic — GSC / GA4 / others

GIG's deck left these cells blank. Filled honestly: the source signal exists but is hand-pulled, and the link from organic activity to leads / sales / GWP doesn't exist because UTM tagging isn't enabled.

Dimension (metrics)	Inside source	Outside source	In Power BI
Product pages, weekly delta (imp, clicks, CTR, position)	Yes	Manual	No
AI-channel report, weekly (sessions → leads, sales, GWP, CVR)	Partial	Manual	No
Blogs, weekly view	Yes	Manual	No
Keywords, weekly view	Yes	Manual	No
Brand vs non-brand	Partial	Manual	No
Product page → funnel tracking	No	No	No

The "sessions → leads, sales, GWP" link is the whole game for organic, and it's not tracked: no UTM rollout means web activity can't be tied to outcomes. Without it, organic attribution is effectively zero.

Internal & offline — the stranded systems

This is where the real damage is. These are GIG-internal and offline sources — the ones that decide profitability and cover half the sales — and they are neither outside their source systems nor in Power BI.

Source	Honest state	In Power BI
Medical (LOB performance)	Not tracked for marketing	No
Email campaigns	No email-level stats at all — the channel is effectively unmeasured	No
Social organic	Largely not tracked	No
SME (segment)	Not tracked	No
Call-centre report	Zero data — roughly half of paid-motor sales close by phone and we have nothing on them	No
Content updates	Not tracked	No
Outdoors / print media (offline)	Not tracked	No
eBao policy data	No access · unlinked — the data that defines profitability	No
Segment, profitability, demographics (Motor)	No access — sits in eBao, never joined to CRM	No
Renewal	Largely not tracked — the largest GWP engine, least visible	No

Competitor — all channels (Organic / Google Ads / Bing / Facebook)

The competitive picture is a clean row of blanks. For every dimension, the honest answer inside is "we don't know," and nothing is outside or in Power BI. We are flying without any view of what rivals are doing.

Dimension	Inside	Outside	In Power BI
Funnel audit (CTA, hero image)	?	No	No
Landing-page review	?	No	No
Videos analysis	?	No	No
Offer analysis	?	No	No
Demographics — age, location, nationality	?	No	No
CTA analysis	?	No	No
Campaign seasonality analysis	?	No	No
Media analysis	?	No	No
Offline advertising	?	No	No
Bids / bid strategy	?	No	No

Read the three tables together and the pattern is unmistakable: in Paid the data is trapped inside the platforms; in Organic it's reachable only by hand; in Internal & Competitor it isn't reachable at all. The Power BI column is "No" almost the whole way down. That column is the entire job, and it's empty.

5 · Why "paid is fine" is a trap

Paid is our best-tracked area — and even there we stop at roughly half of what "good" means. The reason is the one that matters most: we can measure activity, but not profit.

PMax may well be buying policies. It may well be delivering GWP. But without eBao's car-type, demographic and policy-value data joined to it, we cannot tell you whether it is doing so profitably. We could be scaling GWP that loses money and it would look like a win on every dashboard we currently have. GWP is not profit — and the data that separates the two is exactly the data we can't reach.

And it's worse than a missing profit lens. In paid motor — our highest-spend area — roughly half of sales close in the call centre, and we have zero data on them. So even the conversion count is half-blind: we optimise paid motor on the sales we can see, while the other half completes somewhere we can't measure at all.

This is why "paid tracking isn't bad" is misleading comfort. Good-looking paid numbers, with no profit lens and half the motor conversions invisible, is precisely the condition in which you confidently spend more on the wrong things. The ceiling on paid isn't reporting polish; it's the missing join to the profit basis and to the call centre.

6 · It gets worse by channel

Paid is the strong end. Everything else is weaker — and the channel carrying the most premium is the least visible of all.

Paid

~50%

Best-tracked channel. Visible inside the platforms, but no profit join, a manual pipeline, and half of motor sales closing unseen in the call centre. Incomplete, not solved.

Organic

≈ 0%

Tracking is close to nothing beyond hand-pulled GSC. No attribution from organic activity to leads, sales or GWP — UTM isn't enabled.

No stats

No email-level statistics at all. The channel is effectively unmeasured — opens, clicks, conversions, none of it surfaced.

Renewal

Least visible

The biggest GWP engine in the business and the one we can see into the least.

So the honest summary is: incomplete where we're strongest, near-blind everywhere else, and darkest exactly where the money is.

7 · What would close the gap

For completeness — and to be clear this is a GIG-internal unlock, not a tool we're selling — here is the order the gap has to be closed in. Each step is dead without the one before it.

Build the join key. One identity spine linking CRM ↔ eBao ↔ call-centre ↔ web. Until a policy can be traced back to a channel, nothing else is worth doing.
Model the warehouse. A proper semantic layer in Power BI — LOB × channel × country × time — so the data already in the tenant becomes something the team can slice. This is what turns the empty "Power BI" column into "Yes."
Plug the two big holes. Pipe call-centre sales into the spine (recover the missing ~50% of motor) and stand up real organic / UTM tracking (so web activity ties to outcomes).
Bring in the profit basis. Join eBao's car-type / demographic / policy-value data so every view can show profitability, not just GWP.
Then, and only then, value-based bidding. Feed a profit-aware value signal back to the ad platforms. It's last because it depends on all four above.

Compliance note, because it pre-empts the obvious objection: none of this needs data to leave GIG. The join happens inside the M365 / Power BI tenant. Done this way the data-sovereignty story gets stronger, not weaker.

Bottom line

The marketing team isn't underperforming. It's flying with half the instruments missing.

We are at roughly a quarter of where we need to be. The work in flight gets paid reporting to about half — and then stops at a wall that only GIG-internal data access and modelling can get us past. Naming that wall precisely, in our own framework, is the point of this audit.