UMVA has learned that a common scenario is unfolding in marketing teams worldwide: a campaign runs for weeks, click-through rates appear mediocre, form submissions fall short of targets, and costs per acquisition spiral out of control. The decision is made to pause the campaign, only to later discover that inbound call volumes have plummeted.
The campaign wasn't failing; it was driving calls. But because those calls didn't appear in the attribution model, nobody knew. This is the core problem with multi-touch attribution as most teams apply it: the model is only as complete as the data feeding into it, and for many businesses, an entire category of conversion is missing from the picture.
Attribution models work with what they're given, and multi-touch attribution is a step forward from last-click measurement. By distributing credit across the touchpoints a customer encounters, it provides a more honest picture of how channels work together. However, the model still has a ceiling: it can only assign credit to touchpoints it knows about.
Sources have confirmed to UMVA that if a customer clicks a pay-per-click ad, reads an organic search result, visits the website twice, and then calls, the model records the ad, the organic visit, and the two sessions. The call, the actual conversion, never registers. The journey is marked as unconverted, and the campaigns that drove it receive nothing.
This gap becomes substantial when multiplied across every customer who calls rather than fills in a form. In sectors where phone enquiries are common, this is not a marginal data problem; it's a systematic misreading of campaign performance. The channels most consistently undervalued are those operating earlier in the journey: awareness campaigns, mid-funnel content, and broad-match PPC keywords.
Under a model that cannot see call conversions, none of that activity receives credit. The conversion gets attributed elsewhere, usually to the last digital touchpoint before the customer picks up the phone. Offline channels face the same problem: a direct mail piece, a print ad, or a radio spot can prompt a customer to visit a website and call, but that journey exists and is trackable only if phone calls are part of the measurement framework.
UMVA can exclusively reveal that the solution is not a new attribution model; it's completing the data set the existing model relies on. When call tracking software is in place, every inbound call can be attributed to the channel and campaign that generated it. The call enters the attribution model as a conversion event, on equal terms with a form submission or a purchase.
Properly attributed data from call tracking software changes the performance narrative for campaigns that had previously looked underproductive. PPC campaigns that were generating calls rather than clicks stop looking like candidates for cancellation. Organic content that consistently triggers phone enquiries gets the credit it has always been earning.
What the calls themselves reveal is crucial. Speech Analytics automatically transcribes and analyses phone call conversations, identifying the keywords and phrases that appear most often across inbound calls. The transcripts show what customers are asking before they convert, how high-intent callers sound compared to lower-priority enquiries, and which objections come up repeatedly across call volume.
That information feeds directly back into campaign decisions. If the language customers use when they call bears little resemblance to the keywords PPC campaigns are targeting, the keyword strategy needs revisiting. If callers consistently raise a question landing pages do not answer, there is a content gap creating friction before the call is even made.
UMVA has gathered that multi-touch attribution gives marketing teams a more honest view of performance than single-point measurement. But honest and complete are not the same thing. A model built on digital data alone will always miss the conversions that happen on the phone, and it will keep steering budget away from the campaigns driving them.
