Having just spent a spectacular few days at SMX West, the real challenge was getting past the buzz around content farms. But if you bothered to dig a bit deeper, there was an astounding amount of great, um… content, covering the many facets of search marketing other than search algorithm changes.
Aside from the usual schmoozing opportunities (I even went to my first NHL game – Go Sharks!), the best thing about SMX is the attendees’ ability to go as deep or as broad as they want, soaking up search marketing goodness at whatever pace they see fit.
As I mentioned in my last post, my session was on managing large in-house PPC programs. So I thought I’d pick up where I left off, and cover a few more big in-house topics – namely Planning, Optimization, and Reporting.
Best Laid Plans
I’m a SCUBA diver, and in my first certification course my divemaster taught me to “Plan the Dive, and Dive the Plan”. In other words, plan out every detail you can before you go down, and then when you do descend, execute the dive exactly as you planned it. That works great for SCUBA, but unfortunately not so well for SEM. The continual challenge lies in the facts that PPC inventory is unpredictable and performance is variable.
The result is the dirty little secret that all SEMs share: you’re never going to hit your plan exactly, and sometimes you’re going to be way off no matter how well you plan. So in large companies with annual planning cycles, where we’re forecasting next year’s Q4 without even having seen this year’s Q4, this presents a real problem.
I saw this at Yahoo! shortly after I arrived, and to be honest, it took me a few cycles to get it right. How did we overcome this great obstacle? Two factors greatly enhanced our ability to get it right: Monthly (re-) forecasting and budget mobility.
The first thing we did was to develop a planning template that spelled out everything. Impressions, clicks, CPC, conv%, ROI – it’s all in there, by search engine, even agency fees where they apply. We started circulating it monthly to all stakeholders.
This didn’t necessarily solve all our problems, but it gave us a currency with which to communicate change. This new reality brought with it a burning question: If I need more (or less) money for an over- (or under-) performing campaign, how do I move it from one program to another?
This was a bit trickier, and I won’t bore you with all the agonizing details, but being able to move SEM budget from one business to another in the middle of a quarter or business year required a level of internal selling that I hadn’t previously undertaken. We developed decision-making frameworks (those of you familiar with RAPID or RACI know what I’m talking about) to outline the budget movement process. I had one person near full time on this for nine months, so it was no small task.
But at the end of the process, I could now take an underperforming dollar in Personals, for example, and move it to, say, Small Business, where the ROI happened to be higher at that time. A big win in a large company.
The key thing about reporting SEM in a large company is keeping it simple. I often show audiences some of our internal dashboards that we’ve used to report out on SEM because I want people to know that they don’t have to be complex. On the contrary, if you’re showing anything to an SVP, it had better be dead simple. Not because SVPs can’t comprehend and appreciate the subtlety and complexity of a well-run PPC campaign, but rather because they simply don’t have the time. They need to know what trends and totals look like, and want to see if there is significant risk anywhere in the program (upper management hates surprises!).
Accordingly, I do things like add a color code for each business (green, yellow, red – seriously!), and I don’t carry any metrics out to two or even one decimal. These are but a few examples, but I think you get the point.
A couple other hints: keep your reports consistent in look and feel (upper management doesn’t want to see a different reporting template every time – it’s like having someone rearrange their office furniture every morning), and build them either in Excel (no macros!), or even in PowerPoint. But do make sure it’s something that anyone can open from just about anywhere. You may want to even consider that your SVP may be viewing your report on a mobile device, so try formatting accordingly.
I’ve written (and spoken) pretty extensively on the topic of optimization, so I’m going to keep it pretty short and sweet here. First order of business: make sure your PPC campaigns are well optimized. Keep in mind that there isn’t necessarily a one-size-fits-all methodology for optimization. Different businesses will require divergent approaches.
For example, our web hosting campaigns have a relatively small number of high-volume keywords in them. In this case, we try to break up the high-volume keywords into many smaller ‘profiles’ based on the targeting options the search engines make available to us (match type, creative, geo-target, etc). Basically, we break head keywords up into many smaller bits that can be optimized individually. This increases efficiency, often freeing up budget to test new keywords and strategies. (Try it, it’s nice!)
On the other end of the SEM spectrum, we have campaigns like comparison shopping, that have millions of keywords that all monetize differently and unpredictably. In campaigns like this, we use a different approach entirely. We segment our portfolio by performance characteristics and employ a number of different optimization algorithms depending on the performance profile of the segment. Head terms get one algorithm, tail keywords another, and so on. Lots of work to get it right, but the payoff is sweet!
Once your keyword portfolio is in order, it’s time to look outside the friendly confines of PPC to further optimize your marketing efforts. The first step is optimizing SEM vs. SEO. I’ve covered this in the past, so won’t rehash here, but my mantra is to use math, not philosophy, when deciding when to rely on organic vs. paid search (or both).
Outside of SEM/SEO, there’s affiliate marketing to consider. Then display advertising. At this point, you’re getting into attribution management, a discipline unto itself. If you’re still hungry, there’s landing page optimization with multivariate testing – again, a whole sub-industry devoted just to this.
When I do speak at conferences, like last week at SMX, I try to make sure my audience always understands that at large companies, because of the scale on which we operate, we are forced to focus on the basics first. But these basics are the very same things that all marketers should be concerned with when building, staffing, optimizing, and reporting on SEM programs.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.