Over the past few months, I've met with different agencies and their media teams and spoke with them about their planning and buying process and needs. Guess what their biggest complaint was? The answer is obvious -- workflow inefficiency. It is a well-known in our industry that media planners are over-worked and too bothered with overhead and inefficient processes -- leading to sub-optimal media plans and campaign results. But from my perspective, there's also another problem -- a subtle problem that actually might have even a larger negative impact on the media plan and campaign results. I am referring specifically to the lack of available data for media planners to make sound and well-informed decisions during their planning process.
It's no wonder that media planners rely on their own intuition given that there's little data available for the average planner. Here's a classic example: A media planner must develop a consideration set -- a list of the best sites and networks to be considered for the campaign. In most cases, they turn to research tools like comScore, Nielsen, or Quantcast in addition to reusing the same sites from previous campaigns. So, what's wrong with that? Although these research tools can definitely identify which sites are most relevant for the campaign's target audience, they cannot predict or tell how these sites will actually perform. Sure, the planner can pull some data from the last campaign -- but it's not enough to make a calculated decision. So, the planner ends up making an important decision with a huge impact on the bottom line, while seeing only a small portion of the big picture. Not good.
As if making uninformed decisions while creating the consideration set wasn't bad enough, it gets even worse during the negotiation stage. Ask a media planner how much he or she spent for a certain site, section, placement for a recent campaign (or even over the years), and you'll probably get a puzzled stare back. So can the media planner negotiate anything without historical data or any sort of baseline? Not really. The planner just ends up negotiating blindly.
Can you see how absurd this is? We work within a highly innovative and data-driven space that offers advanced algorithms to manage real-time bidding (buying) for non-premium buys and laser-like targeting to match the right message to the right person. Yet when it comes to premium buys -- which still constitute the vast majority of the display ad spend and the core of the media agency work -- critical decisions are still made based on gut feeling.
There is no reason why premium buys cannot be data driven as well. In my opinion, a standard planning procedure should always include few success metrics that will be defined based on historical campaign data. (I would recommend using a primary success metric and one or two secondary success metrics.) These success metrics should be used as a compass to make any decision during the planning process for an upcoming campaign. So what goes into these success metrics?
Success metrics can be performance oriented (clicks, conversions, etc.) or engagement oriented (dwell time, interactions, etc.), and they should be based on historical campaign data. When combined with the reach and site-centric data from research tools like Nielsen and comScore, these success metrics can help the planner focus on relevant sites, prioritization, and building an optimal site list (e.g., consideration set) for the campaign. For example, when Nielsen spits out a report with 50 relevant sites for the campaign's target audience or when comScore includes a list of most-visited sites in a certain site category, the planner should use the success metrics to filter and prioritize these lists. Which of these sites will perform well against the success metrics? What is the expected performance compared to the industry benchmark for the vertical or region? The planner can combine these puzzle pieces together and finally see the bigger picture.
Success metrics can also include cost metrics (e.g., CPM, CPC, CPA), which are obviously very important for any campaign planning. What's less obvious is how to factor in the cost metrics. Too often, pricing is ignored until the negotiations phase or used as the only criteria during the planning stage. Both are bad options. What the planner should be doing is evaluating sites based on a combination of the site's relevancy (using audience composition and site-centric data such as unique visitors) and the site's expected performance and expected price (using historical campaign data).
Historical pricing info is specifically important for the planner during the negotiations phase with different publishers. This is a no-brainer, yet it doesn't happen all too often. The planner needs to evaluate the publisher's suggested price against historical performance or engagement benchmarks and know the average price paid for previous placements and/or packages. The media planner should also not be shy to find out the average price that colleagues at the agency paid. Once planners have this historical data, they need to use it to evaluate the proposal. If the proposal is more expensive than the historical benchmark, planners have two choices: Either go back to the expected performance data (again, focus on your success metrics) and justify the higher price or negotiate the price down. There is no way around it. If planners don't do this, they will just end up negotiating blindly.
To apply all of the above into the planning process, media planners need easy access to the historical data and the appropriate tools to effectively use this data. Data is not only important for audience network buys or exchange buys. It is highly important for premium buys. At the end of the day, most of the online media spend still goes to premium media and will remain so in the near future.
Ronnie Lavi is director of strategic planning at MediaMind.
On Twitter? Follow iMedia Connection at @iMediaTweet.
