Media Mandible

Mainly in that I didn’t take bribes. I mean, I took them, but I didn’t consciously make decisions based on them. In fact, I once had to make a choice between two publishers, and one of the two took me to the US Open. I chose the other publisher, because that was the only way I could convince myself that I hadn’t been bought.

I was also strange in that I understood ad trafficking, behavioral targeting, HTTP requests, Javascript calls, PHP includes, and a lot of other things that people on the business side usually can’t be bothered to learn, because they’re too busy taking each other out to lunch.

I refused to send RFPs with MediaVisor, because, having worked on the publishing side, I knew how annoying it is to be forced to used MediaVisor. A sales rep from Doubleclick once said to me, incredulously, “But MediaVisor will make your job easier!” But, as I told her, my goal was not simply to make my own job easier. I thought there was value in making the jobs of others easier as well… not just clients, but also vendors. People - intelligent people, anyway - work harder when they’re treated with respect and sympathy for their own problems.

However, by “problems,” I do not mean the problem of how to sell me something. I got bored and irritated with sales reps from publishers cold-calling or emailing me and offering to bring in pizza for my whole team, if only we would sit through yet another cookie-cutter PowerPoint that consists of little more than a few screenshots of the site and some data pulled from MRI or Nielsen.

In fact, the little media department I ran produced a templated response to such requests. Emails offering to bring us luch or logo-printed pens would be met with a “mad libs” style document, in which the eager beavers could fill in their targeting methodology, their impressions per month, their rich media specs, and what SOV we might get for a $10k trial campaign. I don’t think anyone ever actually filled it out, but it did get some of the more annoying reps to finally leave us alone.

I was a strange online media buyer, because, sitting at a conference table covered in food and pens and notebooks with the publisher’s logo on them, I would interrup the pretty young rep to tell her that the statistics she was parroting were provably false, that the advanced targeting methodology she was bragging about did not yet exist, that the $10k/day “branding” ad on the homepage was a bad joke.

I was a strange online media buyer, because I spent most of my time thinking about the best way to buy media, and I knew what I was talking about.

I also did some offline (”traditional”) media planning and buying, and there was a lot in that regard that I didn’t know. Reps taught me. A more experienced traditional buyer taught me a lot.

What’s the difference?

So, you know what the client wants, you have explored what happened before you got involved, and you’ve established clear KPIs. Now you’ve got to solidify the budget.
The budget might come to you in any of the following ways:

• You (the media planner) come up with it, and you propose it (internally or to the client).
• The client gives it to you directly.
• Someone in your own company allocates it to you out of a larger budget that also includes creative costs, etc.

Each scenario has its advantages and disadvantages, but it comes down to this. If you’re coming up with the budget yourself, base it on the goals and the KPIs that you’ve already established. Assume an average CPM, an average CTR, a success according to your KPIs (say, a lift in client site traffic of over 10%), and then back into your budget.

Say, for example, that you’re going to run on premium sites those in the CondeNet network (disclaimer: I used to work for Conde), so you need to assume a CPM of about $25. And let’s say that to get the lift, you’re going to need a million new pageviews to the site over the course of a month. And let’s say that, based on a similar campaign, you estimate that, if you get a typical CTR of 0.3%, you’re going to get a conversion of about 1%. (Meaning: Three out of every thousand users will click. One out of every hundred clicking users will buy.) So, given the lowball results you’re predicting, how many impressions will you need to get the results you want? And given the (high) CPM you’re predicting, how much money will it take to buy those impressions?

Of course, this is all somewhat arbitrary. You don’t know the publishers for sure yet, and you don’t know the CPMs until you start negotiating with the publishers. And there’s no sense negotiating until you know how much you have to spend. It’s easy to see the situation as a moebius strip, but it’s frustrating and not productive. At some point, you have to start estimating prices and/or performance in advance, and then fill in the blanks. Make your price estimates high and your performance estimates low, and then do your best to negotiate lower prices and drive better performance than what you predicted.

photo by Jeff Keen on Flickr

So you know what the client wants, and you know what they’ve done before. Now it’s time to establish KPIs (Key Performance Indicators). In other words, you have to determine on what basis the success or failure of the campaign is going to be determined.

The trick here is that the goals the client gives you probably don’t translate directly into realistic KPIs. They might say, to use the most common example, that they want to “increase awareness of the brand.” This sounds good in theory, but it’s hard to measure. Or rather, the way in which it is best measured is up for argument. It’s necessary to have that discussion up front, before the campaign is even planned.

Some common KPIs: click-through rate, ROI, conversion rate, and increase in web traffic.

I’ve seen some agencies talk about “impressions delivered” as if this is a KPI. But really, the number of impressions delivered, once the campaign has been planned, is a measure of basic competency, mostly on the part of the publisher.

If “awareness” is really what you’re going for, then run an awareness study before and after the campaign…if you’re thinking online online only, then it’s common to use Dynamic Logic or someone similar. But that costs money and takes time, and the client may not go for that…which is fine. That gives you the opportunity to say, “Okay, so it looks like we can’t actually measure awareness. So let’s talk about what we are going to measure.”

Clients sometimes squirm during this type of discussion, because they often have no idea what to say. This is a great opportunity for you, the media planner to introduce your own ideas about what the KPIs ought to be. In this way, you are outlining the metrics of your own success.

KPIs should be simple. Draw a distinction in your own mind between what you could use as KPIs and what you should use as KPIs. If you know ad ops and analytics very well, you probably have all kinds of metrics that you could bring into consideration. And maybe you should…later on, as icing on the cake. But for now, keep it simple.

It’s worth noting that the person who drives the up-front discussion of KPIs usually has to be the media planner. That’s because it’s a discussion that takes place around the same time that the deal with the client is being cinched, and during that time, nobody else involved, at either company, is really interested in thinking about or discussing the end of the campaign, or even, in many cases, the details of what the campaign is going to be like. But for your own good (and, it’s worth noting, the good of the client), you have got to make that conversation happen.

Also, don’t fall into the trap of letting the client misuse the KPI conversation to set you up for failure. “Okay, the KPI is clicks, and we expect a 30% CTR,” is not an acceptable conclusion.

Once you know what your client wants, you need to find out what they’ve done before.

Don’t work in a vacuum. It’s surprising how often clients don’t offer this information up front. More often, they’ll come to you and effectively say “Another agency ran some media plans for us that didn’t do very well. We fired them. Maybe you’ll do better.” In fact, sometimes the client may be so eager to leave behind their last partner that they don’t want to talk about who they were or what they did. But you’ve got to get that information. If you can, get hold of the actual media plans and the actual reporting and analysis that the last agency gave to the client.

If the last media planner made obvious mistakes, then you might see your whole media strategy opening up for you right there. Maybe they chose the right sites, but spread the impressions out too thin, or put them in the wrong positions. I had a client who showed me their previous planner’s report, indicating zero clicks and zero conversions over the entire campaign…in other words, they hadn’t been tracking it properly.

Structure sets you free. It’s such a satisfying feeling to be able to go back to the client and say with confidence: “Okay, we know what the other planners did wrong, and that’s where we want to start. Here’s where they screwed up, and here’s what we’re going to change, and here are the results you’re going to see.”

Clients can wrap their heads around that. It’s a lot more appealing than hearing “We ran some reports in MRI, and we cross-referenced the sites that index highest for your target demographic with the sites that offer the lowest CPMs, and we’ve determined that all your customers are waiting for you at Flickr!” Because if that’s all you have, then those customers had better be there, or you’ve just killed your credibility (and that of your tools) in the client’s eyes.

Not that you can’t come up with something beyond what MRI tells you without referring to an old plan…you can, and should…eventually. But old, failed plans are what you’re being compared to anyway, so you might as well seize the opportunity. Heroes are defined by the villains they defeat.

“How do you plan media, anyway?” I am often asked by passersby. Usually, I insist that the asker climb a mountain and sacrifice a goat before receiving the answer. But if you find this post, it’s free.

This is (not coincidentally) analogous to a lot of sweepstakes promotions I’ve been involved in, wherein you have the choice between jumping through hoops to enter the sweepstakes, or just entering it. Some people like to jump through hoops. Some people think that life is hard, so given two options, hard and easy, they choose hard most of the time. It just feels more…honest, I guess.

I am not one of those people, though.

Step One: Find out what the client wants.

You’d think they would just tell you, right? But that’s often not the case. A surprising amount of the time, the client (or an account manager) will come to you with a mandate like “Onceler Corp. wants a media plan for their ‘thneeds’ brand by Friday. It should include an email and, uh…animation.” It is then the media planner’s job to push back and ask (among other things): What does the client want?

Sure, they want a media plan, but why? Are they trying to get people to click on a banner and immediately order thneeds online (direct marketing), or are they trying to get across the idea that thneeds are things that everyone needs (branding)? Why have they decided to run ads at this particular time? Are they launching a new product? Doing damage control on one that exists? Making a show of force to counter a competitor’s advertising? Or did an executive at the client company just announce one day “We should really be doing more of that…what do they call it…online advertising” (this happens more than you think)?

Getting these questions answered is more important than anything else at this point. And this is crucial: It’s more important than delivering a media plan. Because if you give in and just make something up, then before you know it, you’re going to have to implement it and then defend the results. Fortunately for you, asserting that you need these questions answered before proceeding is usually an effective pushback. Either the client realizes they don’t know what they want, or the account manager realizes they don’t know what they client wants. In either case, they decide to mull it over (or you pro-actively help them figure it out, if you want to swim in those waters), and you live to plan another day.

Of course, the best way to avoid this problem is to make sure you’re involved in the account from the start, i.e. even before there is talk of planning media, so that you have all the relevant background information. I do not agree with Seth Godin’s implication that meetings are usually a waste of time. Sure, they often don’t accomplish what they set out to accomplish. But for the observer who is simply interested in soaking up the nature of the situation under discussion, filing it away for future reference, meetings (other people’s meetings, that is) can be a great resource.