Here's how CPL needs to change

Agencies are called in and asked to justify their existence, which, millennia of human inquiry notwithstanding, is reduced to ROI. While ROI seems to hold within its syllables the key to doing capitalism well it actually deafens us to ignore other metrics like quality and attribution.

Publishers then accept CPL deals, of which some are eye-wateringly bad. At a simple level the payout sounds good and risk and reward is factored into the equation. At other times these are so bad that I’ve seen R1 being offered for a vehicle insurance lead (#TrueStory).

At the end of the campaign publishers and agencies tally up reports. Too often the brand’s CRM software is not fully integrated so attribution becomes murky toward the end. Many publishers I know actually go through the conversion process themselves or by proxy, and later expense the purchase, in order to provide their own check to the success report coming in.

Usually a bucketload of impressions has been thrown overboard in order to reel in the leads and the CTR comes back with an underperformance of normal campaigns, which then tends toward zero against the leads.

Here are the problems

If media has been allocated, campaign runs to term and leads have not been generated - all it reveals is that everyone has done their job at the lowest box-ticking level. The media owner has incurred costs of the inventory, the agency has generated effort in running the campaign and the brand has defensively not spent a cent except in the unlikely case of a payout. The only one who hasn’t kept up their part of the implicit process is the targeted user.

Optimisation, basic yield management done by ad ops, will require work on the sell-side but at low real values it will not generate any urgency. Where yield management sparkles is when publishers will target against their own first party data, or even when other data is brought into play. But, as we know, investment into Data Management is hampered by lack of cash for the many and lack of foresight by the few.

Brands get exposure for every valid impression but this is not being reflected. If the creative is selling me a new car with a call to sign up for a test drive, why is the marketing of the new car not being allocated to a line-item?

Here’s the solution

The pricing model has to be made more robust. A base CPM has to accompany every CPL campaign to reflect the marketing that is actually going on. This will also cover the incentive to optimise.

Other media channels would never dream of giving away media without it being expensed. Ours is the most measurable and effective so giving it away at below cost is an exercise in making digital marketing look stupid.

The entire value chain needs to know the lifetime value of a lead, average order size, qualified lead rate and conversion rate.. Sometimes this information is available anecdotally, but it can also be found going through a company’s financial statements or luring an actuary away from a cubicle or saddle with offers of wild merry-making.

There is a place for Cost Per Lead. That place is where audience data will come into its own. But, as publishers have known for years, easy come easy go - a low Cost of Campaign is hard to climb back from. The same will happen to agencies who return limited lacklustre leads at low cost, which brands will come to expect. When we find we can’t improve on this by augmenting with first party data then we’ll find that ourselves running campaigns in a shallow pool of low quality leads at low propensities to purchase. With red water. And a big shark and even bigger croc.