Attribution
Marketing Attribution
What is attribution?
- Splitting conversion credit across multiple touchpoints
- Looking only at “last click” misses early contributions
- No model is perfect, so understanding its limits matters
The first thing that changes when you adopt attribution is not the look of the report but the unit discussed in budget meetings. Channel-total conversations like “300 leads this month, KRW 50 million (≈ US$37,400) in ad spend” give way to journey-level conversations like “in what order did touchpoints build up to this one contract?” It also settles the situation where the ad team and the content team each claimed 100% of the credit for the same contract.
Models fall broadly into rule-based and data-driven types, but in B2B there are conditions to solve before choosing a model. Evaluation periods stretch from three months to a year, and a single purchase involves 3–7 people, from practitioners and team leads to procurement and executives, so tracking one person’s browser cookies will inevitably break partway. That is why you group touchpoints at the account (company) level rather than the individual level, and tally contribution up to lead creation separately from contribution up to pipeline and revenue, instead of mixing them in one table. Marketing mix modeling (MMM), which estimates channel effects from weekly aggregates without tracking individuals, serves a different purpose, so it is better not to treat it as a substitute for attribution.
In practice, the most common mistake is reading attribution numbers as proof of causation. Attribution is just a ledger that splits credit for conversions that already happened according to fixed rules; it does not tell you whether revenue would actually fall if you turned that ad off. As a result, budget keeps flowing to clicks from people who searched your company name, while offline seminars, industry referrals, and community mentions that can’t be tracked are recorded as zero contribution and cut as “non-performing channels”. Zero contribution does not mean no contribution; it means not measured, and channel contributions often add up to more than actual revenue because of double counting. Once a quarter, validate the ledger with an experiment that deliberately scales back a specific channel.
John KimInstructor John Kim, author of B2B Marketing Design and CEO of Nanum Consulting
Euijung KimBrand strategist with 30 years in advertising, formerly a division head at Daehong CommunicationsThe concept took hold alongside the growth of digital advertising and web analytics (such as Google Analytics), out of the need to measure the contribution of multiple channels.
Consider a hypothetical example. Company A, a B2B SaaS firm with 210 employees and annual revenue of KRW 48 (≈ US$1) billion, was spending KRW 62 million (≈ US$46,400) of its KRW 80 million (≈ US$59,900) monthly marketing budget on search ads. The last-click report showed that search ads drove 82% of contracts. Yet quarterly contract volume had stayed flat at around 40 for three quarters in a row.
- The marketing lead joined the CRM’s inquiry history with web analytics sessions by company domain and rebuilt the touchpoint history for each of the 118 contracts signed in the past 12 months, one row per contract.
- The rebuilt data showed an average of 9.4 touchpoints per contract and a median of 147 days from first touch to contract, with search ads mostly appearing after the eighth touchpoint.
- The sales director pointed out, "Aren’t those people just typing our company name?" When brand keyword traffic was separated out, it accounted for 61% of search ad clicks.
- Contracts whose first touchpoint was a webinar or an industry report download came to 41 deals, 35% of the total, yet under the last-click model their attribution was recorded as zero.
- They ran a six-week incrementality test that cut brand keyword ad spend in half, found that traffic dropped by only 7%, and moved the remaining budget into two webinars a month.
- Two quarters later, the monthly budget was still KRW 80 (≈ US$1) million, but contracts rose from 40 to 57 per quarter, and acquisition cost per contract fell from KRW 6.1 million (≈ US$4,560) to KRW 4.3 million (≈ US$3,220).