How people reach Lucy's, how we could capture where they came from, and what Google and the marketing report can see at each of three conversions. The capture steps are suggestions for Jess to shape around how she already works.
The counts above are of the 19 steps on the flow diagram.
The risk: Google can see form fills, but almost none of the consignors those fills become. If the ads are judged on what we can currently see, they will look weaker than they are, and spend could be cut for the wrong reason.
Recommendation: hold any decision to cut or pause seller ads until the click-ID data has had time to mature. Email enquiries take three to five weeks to sign, so a fair first read is mid-December, and a full 90-day read is early February 2027.
Three views of the same facts. The flow diagram shows how people arrive and where attribution is lost. The step list gives the detail on each step. The timeline shows when the data can be trusted.
One flow, for Google Ads and sellers. The capture steps shown are suggestions for Jess to adapt to her own process. It has three conversions. Each one is measured twice: once for Google, so it can learn, and once for the marketing report, so we can see it ourselves.
On a narrow screen, scroll the diagram sideways.
What Google can learn from. Only a lead with a click ID, or a match to an earlier web lead by hashed email or phone, can be credited to a Google ad. Someone who rings and says "I saw a Google ad" counts in the marketing report but, on Google's published guidance, cannot be credited in Google. Brandunity will confirm this with Google before relying on it.
One card for each box on the flow diagram, in the same order. Click a card for its evidence, its numbered drop-off points and the next move.
When the data is, and is not, trustworthy. The dashed line is today. A fair decision on the ads sits to the right of it.
Bands are dated from the tracking record and Jess's answers. The two read dates are Brandunity's recommendation, based on three to five weeks to sign and Google's 90-day window.
Every enquiry becomes one row in Jess's tracker. These are the fields we suggest each row carries, so the marketing team can attribute enquiries and measure the return on spend. How they are captured is for Jess to decide. Switch to see what is captured today.
Referral is not a separate way in. A referred person still arrives by form, phone, DM or a trip. A "Referred by" field on that row could hold it.
Two questions, two fields. How they contacted us (a phone call) is not how they first heard of us (an Instagram ad). Capturing both lets a phone call be traced back to an Instagram story.
A suggested pick list for how they first heard:
A suggested question:
These are limits of the channels and the technology, not faults to fix. Where we cannot track, asking and noting the answer is the alternative.
| Gap | Why | What could help instead |
|---|---|---|
| Instagram DMs and stories | No click, so no click ID. | Asking how they heard, noting it in the tracker, and reporting Instagram as its own channel. |
| Safari visitors who return after 7 days | Safari caps script cookies at 7 days. | Hashed email and phone (enhanced conversions for leads) match them instead. |
| Private windows, ad blockers | The cookie or script can be blocked. | Accept a floor of lost attribution; compare Google's count with the form export. |
| Sales more than 90 days after the click | Google rejects older uploads. | Use the early milestone, the signed agreement, as the signal. |
| Word of mouth, referrals, a friend's story | No digital trail. | Asking how they heard and noting a "Referred by" name on the row. |
| Who saw an ad, then enquired another way | View-through across channels is not visible. | Judge by direction of travel, not by single-channel counts. |
These are suggestions. Jess knows the process best and can shape each one.
Question for Lucy's: are any Instagram or Meta ads running for sellers? This map assumes none. If there are, they follow the same flow with different IDs.