When the time comes to review a media plan, the report usually arrives with a cost column and several versions of what was a ‘result’. The paid search presents requests. The video displays views. The external advertising presents estimated contacts. To redistribute the budget, start with the decision you need to make: increase reach in a specific audience, generate more commercial opportunities, or sell with margin.

Give each channel a metric that matches its purpose

If the video serves to present a brand to new buyers, it makes sense to measure the reach in the chosen audience and the average frequency. If the paid research serves to capture budget requests, it matters how many of these requests reached the commercial team, met the combined criteria and became opportunities. IAB Europe distinguishes efficiency from delivery, brand and salesIt’s a way of knowing what results can be compared.

O release of LinkedIn television ads in the UK illustrates the difference. In a Cisco test, the range result indicates how many medium-sized companies television has added to the audience of the videos in the feed. The form fillout rate corresponds to the contact ads that followed in the feed. Each number responds to a different decision on the purchase of media.

Put costs and results on the same basis

For each campaign, record the dates, the public, the geographical area and the announcement that has been served. Please indicate whether the investment includes only advertising space or also production, fees and tools. For enquiries, combine a cut-off date that gives the commercial team time to evaluate them. A campaign completed yesterday may seem worse than another whose requests have been worked for a month.

In a hypothetical example, two campaigns spend EUR 2,400 each. The first generates 80 enquiries, of which 12 are qualified. The second generates 45 applications and 18 qualified. The first channel charges, on average, EUR 30 per order. The second is for the 133 euros per qualified request, compared to the 200 euros of the first. The order changes when measuring the result that the sales team actually uses. If the budget includes production on one channel and only buys space on the other, the comparison still remains to be corrected.

The same care applies to audience metrics. Adding video and street range can count twice who had contact with the two. When there is no overlay estimate, indicate the channel range and explain the limitation of the total. Web Summit’s external advertising case it also shows why the rotation of ads on a screen and the number of people reached need different sources.

O Google Analytics campaign URL guide explains how to use the parameters utm_source, utm_medium and utm_campaign to identify the origin, the medium and the campaign. Use the same convention on all links and test the route to the sending of a form. If the CRM stores the origin with the request, the commercial team can link that record to the qualification and the possible sale.

Choose in advance how to count duplicate contacts. A buyer can click on an ad, fill out a form and return by email two days later. It is a person with two registered contacts. If the goal is to capture new companies, it may still be necessary to bring together different people from the same organization. Documenting these rules allows you to compare reports from different months without switching units in half.

To compare options across paid search, display and social, see the guide to PPC platforms.

Reserve part of the budget to test the difference

The volume that a platform assigns to the campaign helps to follow the delivery and adjust ads. To decide whether to strengthen the investment, it is worth asking how many sales would have happened during the same period anyway. guidance from IAB and IAB Europe on incremental measurement describe ways to build a comparison, including control groups.

A chain of stores can test a campaign in comparable markets, keeping the rest of the plan and promotions as similar as possible. Before starting, you need to choose which sale will count, for how long and what difference would justify repeating the investment. If the sample is small or the stores have very different movements, the result can serve to prepare a better test instead of sustaining a budget transfer.

At the review meeting, record what changes and what data returns to the table. For example, keep the budget that brings qualified requests and test whether a new channel adds coverage for the next 30 days. At the end of that period, the team will have a concrete decision to make with the data they agreed to collect.