The Full Journey: TV Ad → Google Search → Website → Phone Call → Revenue
Google doesn’t motivate a search. It enables it.
Nobody randomly types a brand name into a search bar. Something started the journey — a TV ad, a radio spot, a billboard, a conversation, a change in the weather. The search is a response to a stimulus. Google captured the action. Something else caused it.
I’ve been saying this since my first Forbes piece in 2019, and the data has only gotten louder since then. Google’s own research with Kantar found a positive uplift in search queries for every single one of 98 TV campaigns analyzed. Every one. TV doesn’t compete with search. TV creates the searches that Google then monetizes.
But here’s the problem: most measurement systems credit Google for the conversion and stop there. The stimulus that started the journey — the TV ad that put the brand in someone’s head — gets nothing. Because nobody traced the chain.
The chain nobody traces
Let me walk you through what actually happens when advertising works. Not in theory. In data.
One ad, one customer, one chain
A TV ad airs during the evening news in Minneapolis at 6:22 PM. The ad is for a home services company — HVAC, plumbing, that kind of work.
A viewer in the Minneapolis DMA sees the ad. They don’t act immediately — they’re watching the news. About two and a half minutes later, they pick up their phone and search the company name on Google.
Google records the search. The viewer clicks an organic result. Google assigns a click ID — a GCLID — that ties this specific click to this specific search in this specific market at this specific time.
The viewer lands on the company’s website. The website’s tracking tag fires and records the session — device, location, referral source, landing page, timestamp. The session is tied to the GCLID.
The viewer browses for a minute, finds the phone number, and calls. The call tracking system records the call — duration, whether it was answered, whether it’s a first-time caller, the phone number. The call is tied to the web session through the GCLID.
The call becomes a booked appointment. The appointment becomes a job. The job becomes an invoice in the CRM — $4,200 in revenue.
That’s the full journey: TV ad → search → website → phone call → booked job → revenue. One chain. One customer. Real money.
Google gets the credit, TV gets nothing
And in most measurement systems, Google gets 100% of the credit. The TV ad that started the entire sequence gets zero.
Why the chain breaks
The chain breaks because each step lives in a different system with a different time reference, a different identity model, and a different owner.
The TV ad airing lives in an ad detection system. It uses broadcast time — which runs 6 AM to 6 AM and is timezone-naive to the local market. The web session lives in analytics — UTC timestamps. The phone call lives in a call tracking platform — local time. The CRM data lives in ServiceTitan or Salesforce — business timezone. The Google Ads click lives in Google’s system — account timezone, and critically, Google only provides a date for each GCLID, not a timestamp.
That last one matters. When your attribution window is measured in minutes and the source platform only gives you a date, you’re missing the precision you need to connect the dots. You know a click happened on Tuesday. You don’t know if it happened at 6:24 PM — two minutes after the TV ad — or at 11 AM, eight hours before.
Most systems don’t even try to connect these data sources. The TV vendor measures TV. Google measures Google. The call tracking platform measures calls. The CRM measures revenue. Each system grades its own homework. Nobody shows you the chain.
The IAB’s 2026 State of Data report confirms this: 75% of buy-side leaders say core measurement approaches are underperforming. The data exists to trace the full journey. The infrastructure to connect it hasn’t existed — until recently.
What it takes to trace it
Tracing the full journey requires solving three problems that most measurement platforms avoid because they’re genuinely hard.
Problem 1: Time reconciliation. Every data source uses a different clock. You have to convert everything to UTC for ordering, account for broadcast time conventions, handle daylight saving transitions across markets, and reconstruct broadcast time for reporting. A three-minute error in time conversion shifts your attribution from the peak of genuine response to near-zero. We maintain our own time source with microsecond precision specifically because this problem is that sensitive.
Identity and geography
Problem 2: Identity resolution. The person who saw the TV ad, the device that searched, the session that visited, the phone that called, and the customer record in the CRM — these are all different identifiers in different systems. Session IDs, click IDs, device fingerprints, phone numbers, CRM contact records. Connecting them requires deterministic matching where possible and scored inference where necessary. Not probabilistic guessing — layered identity resolution that starts with what it can confirm.
Problem 3: Geographic verification. The TV ad only aired in certain markets. The web session came from a specific location. The phone call originated from a specific area. If the ad didn’t air in the market where the response occurred, the chain is broken — there’s no causal link. This requires knowing the actual delivery footprint of every publisher (not a DMA approximation, but the real broadcast polygon), resolving every response to its geographic origin at the zip level, and comparing the two.
These aren’t nice-to-have features. They’re the physics of proving that one event influenced another. Without time precision, you connect the wrong events. Without identity resolution, you can’t follow the person. Without geographic verification, you credit ads that never reached the responder.
The GCLID precision gap
There’s also a fourth problem that nobody talks about: the GCLID precision gap. When a viewer sees a TV ad and searches on Google, the click generates a GCLID. Google provides the date of that click — but not the time. For a system where attribution windows are measured in minutes and the response curve peaks at 150 seconds, a date is not sufficient. You need to know whether the click happened at 6:24 PM (two minutes after the ad) or at 10 AM (eight hours before). Because our system operates its own first-party tracking tag, when a GCLID appears in our tracking data, we add the precision timestamp from the tag collector. That triangulation between Google’s click record and our own tracking data gives us the second-level precision that Google doesn’t provide. It’s a small technical detail that changes whether the attribution is real or noise.
The 64% that Google captures but doesn’t cause
Here’s a number from our data that I think about constantly: 64% of TV-attributed web sessions arrive through Google Organic search. Not paid search. Organic.
That means when someone sees a TV ad and responds, nearly two-thirds of the time they go to Google and search — they don’t type the URL directly. Google captures the intent. TV created it.
In a standard analytics setup, those sessions show up as “Google / organic” in the source report. The natural conclusion: Google is driving 64% of our conversions. The budget recommendation: spend more on SEO, maybe increase paid search bids.
But the truth is the opposite. TV created the search intent. Google was the mechanism, not the motivation. Cutting TV to fund search would be cutting the stimulus that creates the demand Google captures.
You can’t see this in Google Analytics. You can’t see it in your TV vendor’s report. You can only see it when you trace the full chain — from the ad airing in a specific market at a specific time, through the search that followed, to the session, the call, and the revenue.
Closing the loop makes everything work better
The full journey doesn’t just explain what happened. It makes what happens next work better.
When you can trace from TV ad to revenue, you can upload that revenue-linked conversion back to Google Ads. Not just “a call happened.” Not just “a form was filled.” But “this specific click led to a phone call that became a $4,200 job.”
That changes how Smart Bidding works. Google’s algorithm learns that certain clicks — from certain searches, in certain markets, at certain times — lead to actual revenue. It starts optimizing for revenue instead of clicks. The bidding gets smarter because the training data reflects reality.
The same principle applies to Facebook’s Conversion API, to programmatic platforms, to every buying engine that uses machine learning. The quality of the optimization depends entirely on the quality of the conversion signal. The IAB’s CTV conversion API guidelines were built on exactly this insight — that closing the outcome gap requires sending real business results back to platforms. Two-thirds of advertisers who implemented CTV CAPI reported improved ROAS.
Most advertisers send Google a pixel fire. We send it the revenue.
The compounding effect
The compounding effect is real. When Smart Bidding learns from revenue data instead of click data, it shifts budget toward the searches, markets, and times of day that produce real business outcomes. Over weeks, the algorithm gets sharper. Cost per acquisition drops — not because you gamed the bidding, but because you gave the platform the truth about what works. Your TV buy becomes more efficient because the digital platforms it feeds are learning from better data. Your digital becomes more efficient because it’s optimizing for revenue, not vanity metrics. The whole system improves because every piece is learning from reality.
The industry is finally asking for this
Performance TV became the number-one investment channel in 2026, accounting for 24% of total media spend. CTV ad spending is projected at $38 billion, up 14% year over year. Netflix and Comcast launched Conversion APIs. The IAB published CTV conversion guidelines. AdExchanger called it “the year CTV goes full funnel.”
The entire industry wants TV — linear and streaming — to prove itself like digital does. To show the full chain from ad to revenue. To close the loop.
Here’s what nobody’s saying out loud: this is harder for CTV than for linear TV. CTV has the advantage of IP-based matching, but it has the disadvantage of fragmented identity across apps, devices, and households. Linear TV has the disadvantage of no IP targeting, but it has the advantage of geographic determinism — you know exactly which market received the ad because broadcast delivery footprints are physical, not probabilistic.
We’ve been tracing the full journey for linear TV for years. TV ad detection → search attribution → web session → phone call → CRM revenue. The chain that CTV platforms are just now trying to build, we’ve already built for the harder signal.
Start with the question nobody’s answering
The next time someone shows you a report that says “Google drove 64% of conversions,” ask one question: What drove the Google search?
If nobody can answer that, you’re looking at a fragment of the journey and calling it the whole picture. You’re crediting the mechanism and ignoring the motivation. You’re optimizing the middle of the funnel while starving the top.
The budget trap
This is happening at scale right now. 68% of multi-touch attribution models over-credited digital channels by 30% or more. TV budgets are being cut to fund digital — because digital can “prove” it works and TV can’t. But TV is the stimulus that creates the digital activity. Cut the TV budget and the Google performance drops. Nobody connects the two because nobody traces the chain.
Ad Age says AI won’t fix measurement if the foundation is broken. They’re right. But the foundation isn’t the algorithm — it’s the refusal to connect the data sources that show you the full journey.
The full journey is traceable
The full journey is traceable. It requires time precision, identity resolution, and geographic verification. It requires connecting systems that were never designed to talk to each other. It requires asking a harder question than “which channel gets credit.”
But when you trace it — when you follow a single customer from the TV ad they saw to the search they typed to the website they visited to the phone call they made to the revenue they generated — you see something no vendor report shows you: how your channels actually work together.
That’s what the Insights & Data Engine was built to do. Not credit a channel. Trace the chain. The methodology is published. The chain is visible in your own data.
Related: Your Google Ads algorithm is only as smart as the data you feed it →