Австралийские маркетологи теряют бюджеты из-за искаженной атрибуции и устаревших метрик

Most Australian marketers are not short of data. Based on two conversations I came across in recent weeks, the problem is they are optimising for the wrong number, writes Ben Henzell, founder of digital marketing agency BFJ. A Sydney industry roundtable for marketing leaders (hosted by NASDAQ-listed martech outfit Quantcast), included teams from Swyftx and Moomoo, to talk through the challenges facing the sector. Attribution came up immediately, and the frustration revealed that marketers are not short of data, yet the data they work with, while technically accurate, is still practically misleading. The version of this problem I hear most often goes like this…. A prospective customer sees a brand across a run of premium environments over several weeks – catch-up TV, digital audio, a billboard on the commute home. None of that exposure, that’s building credibility, registers as a conversion. Yet when that person searches for a brand option in their category or clicks a retargeting ad, Meta or Google takes the credit. Last-click attribution did not invent this problem, however it does help enshrine it. One marketer in the room described their approach to measurement as essentially last-click, because that was the story their CFO could follow. That is a rational response to an institutional constraint, but it has a consequence: spend concentrates in the channels with the cleanest attribution trail, which are rarely the channels doing the heaviest lifting earlier in the purchase decision. Michael Engeman, CEO of Ainslie Group exemplifies that point: “We know from our own data that a customer who eventually opens an account with us has usually been in-market for weeks, sometimes months. Yet our reporting attributes Google search or a Meta ad, as if nothing else happened. That’s clearly not the full picture, and if we only funded what that picture shows, we’d be spending less on the very channels that are building the trust our category depends on.” The irony is that the problem compounds with customer quality, because for fintechs, high-value investors do not make impulse decisions. Instead, they research for weeks across a wide range of environments, and the consideration cycle from first exposure to eventual conversion is long. The longer that cycle, the greater the gap between the channel that built the brand case and the channel that closes the conversion, and the more systematically misleading a last-click model becomes. Fintech is chasing high-long-term-value customers using a measurement architecture designed for impulse purchases. Les Binet’s research puts a number on the broader version of this problem. Budget, he argues, is nine times more important than ROI as a driver of profit, yet most brands make decisions as though the inverse were true. For specialist sectors, the stakes may be even greater: trust is the primary barrier to conversion, which means brand investment is the part laying the foundations in the minds’ of your customers that you’re a trust-worthy brand. Vital for new category entrants. The second challenge that came up at the table is connected. Many marketers often have a strong instinct about where their audience lives online. That instinct tends to produce a list: a set of approved placements, known environments if you will, that feel like familiar channels. The brief goes out, and the buy follows the list. The problem with the list is that it is a prediction, and audiences rarely behave predictably. A high-value investor researching their options might be on a personal finance blog at 7 am, on a gaming platform at lunch, and on a news aggregator in the evening. That person is reachable across all three. The placement list finds them in none of them, because none of them was on the brief. One example from our own experience that illustrates the point – a campaign run for a retirement home group. The audience (55-plus decision-makers) turned out to be reachable through mobile gaming inventory. That insight did not come from a planning session but from following the audience data to where the audience actually was. The placement list would never have included it, and the brand would never have known. For any brand, whose ideal customer is high-value and in a long consideration cycle, the cost of that instinct and trust compounds. The brands that show up across the full research path (and in the environments their audience actually inhabits) are the ones that earn the conversion. Attribution models that only see the last step cannot tell you whether that work was done. What connects both problems is the same underlying pressure: the requirement to prove return before committing spend. These marketers are not making irrational decisions; they are making the best decisions their measurement frameworks allow. But those frameworks are calibrated to capture what happens last, and often the customers with the highest life-time value are the ones with the longest journeys to conversion. Until measurement can connect brand exposure to eventual acquisition, the channels that build consideration will continue to lose budget to those that claim credit for it. Binet calls it a death spiral. In fintech and more, it is spiraling through the media plans of some of the most data-fluent marketers in the country.