How many contacts does a customer need before buying? A new study makes the question tempting, but a count of remembered contacts cannot tell a marketing team how many ads to buy. A better operating question is which evidence supports each claim about the journey and which decision that evidence can justify.
Read the new research as a description
In a release dated 29 October, IFH MEDIA ANALYTICS and MEDIA CENTRAL describe interviews with 1,000 consumers about purchases in food, DIY, electronics and furniture. The research reports six channels on average. Respondents who described purchases as spontaneous and advertising-induced recalled 43 advertising and information contacts, compared with 19 for need-driven purchases with active search.
These are recalled contacts, not an experimentally established frequency target. They include information as well as advertising. The German study covers four sectors; the findings are not a universal benchmark for every country, product or decision window. The public release also comes from organizations involved in the commercial marketing market.
Keep three kinds of evidence separate
Our proposed framework has three layers. First, recollection: interviews or surveys describe what customers remember and how they explain their choice. Use this to form hypotheses about missing influences. Do not turn the share mentioning a channel into its share of revenue.
Second, observation: orders, CRM records and consistently collected channel events describe recorded actions. Define the unit, time window and missing coverage. A visit, an identified customer and a purchase are different units. A measurement system that joins some of them still needs an explicit account of what it cannot observe.
Third, causality: a suitable controlled test asks what changed because of an intervention. A holdout, where feasible, can challenge the budget hypothesis. Its design must fit the market, sample and spillover risk. A sequence of observed contacts alone does not provide the same answer.
Build a measurement brief around a decision
Start with one product category and one purchase situation. Keep a planned furniture purchase separate from an urgent household replenishment. Define a decision window before analysis rather than choosing one that makes the preferred channel look strongest.
- Decision: name the budget or journey change under consideration.
- Hypothesis: state the customer task the channel is expected to help.
- Evidence: label each measure as recalled, observed or causal.
- Blind spots: record missing offline contacts, identity gaps and survey limitations.
- Contradiction: specify what result would make the team reject its hypothesis.
For example, imagine a furniture retailer considering a change to catalogue distribution. Customer recollection can suggest a role in discovery; site behavior can describe subsequent product exploration. Neither establishes the extra sales caused by the catalogue. A feasible controlled comparison, with a common sales definition and agreed evaluation window, is the next question. This is an illustrative decision process, not a result reported by the study.
Use contact counts to investigate, not prescribe
If customers recall many contacts but recorded paths look short, investigate missing coverage before declaring one channel dominant. If surveys suggest discovery value but a controlled test finds no commercial difference, inspect the design and uncertainty rather than choosing whichever number supports the existing plan.
Keep frequency management as a separate question. Repeated ad impressions, encounters with different channels and remembered information are not interchangeable units. There is no sound route from the headline number 43 to a campaign cap of 43 impressions.
For the next review, bring one page that separates what customers say, what systems record and what the business has tested. That distinction makes uncertainty visible and gives the team a practical way to improve evidence before reallocating money.
