Industrial Strength Search Marketing (SEM, SEO)

This month’s column for Search Engine Land talks about SEO at Yahoo! and how we’ve been forced to re-think our approach to implementation. It goes something like this:

Do you drive to work? Do you take the same route every day? What if one day, while you were driving on your normal route to work, you came upon a detour that forced you to take a different route, one you might have known about but never bothered to try. Funny thing is, it turns out this is actually a better way to get to work, and from that point on you start taking the ‘detour’ instead of your normal route, and it becomes your everyday path to work.

This happened to us with SEO and how we deploy it at Yahoo. We took the approach that most SEOs end up using, doing site audits and making recommendations. It wasn’t working and we were forced to change our strategy. Of course, it turns out that our new strategy was the right approach all along….

If You Build It SEO-Friendly, They Will Come

The Story

“So I audited the website, made all kinds of great SEO recommendations, and nothing. None of them got implemented. Not a single one.”

Sound familiar? This story is told all too often, at search industry events, cocktail parties, and around water coolers at web-based companies everywhere. It’s a story that starts with a VP hiring either a Search Engine Optimization (SEO) manager or agency, continues with the company spending a significant amount of rsources to generate detailed recommendations to enhance the woeful state of SEO at the company’s website, and ends with the inability to close on these recommendations for one or more of a dozen reasons that now seem obvious through the 20/20 lens of hindsight.

Nobody Left to Mind the Store

Often at the root of this problem is a lack of complete buy-in from all stakeholders in an organization, and frequently it’s either the CEO or the IT/web dev group (CTO or equivalent) who’s not on board. However, at large, multi-product companies like Yahoo!, there is another dynamic we discovered that routinely derails even the most well thought out SEO plans. Especially at places like Yahoo!, where there is always a large number of concurrent initiatives within each property competing for a limited pool of engineering resources, it’s very challenging to actually find the bodies necessary to implement SEO recommendations, even if the recommendations are perfectly detailed and correctly prioritized. It’s not because the proposed changes are not valuable or necessary, it’s because the product team that built the site in the first place has since been disbanded, and its members are now part of other teams working on other projects. Simply put, there’s noboby left to implement significant site changes. Bug fixes, sure. Mod re-writes, forget about it.

Stop The Insanity

We all know deep down that it is much cheaper and easier to build SEO into a development project, rather than to release it and retrofit it later. While this is a basic tenet of SEO, all of us routinely ignore it, building and releasing SEO-unfriendly sites, promising to go back and fix them later. I’ve done it, you’ve done it, we’ve all done it. Some of you are doing it right now. Well, folks, it’s time to stop doing it.

Because Yahoo.com is one of the biggest sites on the ‘net, the economics of this dilemma are magnified to such a degree that forces us to behave differently. At some point last year, after repeated attempts to make and implement SEO recommendations on some of our larger properties, we decided that we were no longer going to fix what was broken. Rather, we made a very conscious choice to focus only on what is in development, ignoring the myriad examples of disastrous SEO practices out there on our properties, choosing to replace them slowly over time with new, SEO-friendly product. We made this decision not because we’re disciplined or we have the bottom line in mind, but rather because of the resource-constrained nature of product development and the immense challenges it poses for SEO.

Making SEO Work for You

Once we made this decision, it became obvious that we needed to build SEO into the product development cycle. I know of several people in our industry who figured this out a few years ago and built a business on it. The same rules apply here to us, and probably to you, too. Some very smart folks in our company really dug into the development process and identified several critical touchpoints in the product development cycle where SEO goodness needs to be assessed and addressed.

The SEO touchpoints for you, your company, and your development cycle will vary, depending on the scale and scope of your efforts. We identified six touchpoints on our end, from product conception through post-launch, and developed customized SEO checklists to use at every stage. These checklists have our SEO managers evaluating SEO hotness at each stage in the development process. That means we have dedicated SEOs reviewing requirements documents and wireframes, and filing bugs with the engineering teams when things don’t look right. This is the time to file bugs, when engineers are working on the product, not after it’s released. This approach aligns and integrates SEO with product development, the goal being to launch product that is already SEO friendly. This creates a mountain of work up front, but it will greatly enhance your chances of success moving forward, and over time your work will become much more efficient, and ultimately more effective. It takes patience, committment, and perhaps most importantly, support from your executives.

If you try this approach, and I absolutely urge you to do so, I guarantee you will find product managers who will tell you things like “SEO is all well and good, but I’m not holding up my product launch for an SEO feature”. To which you can now say, “If you follow my process you won’t have to delay your launch for SEO, because it will be built right into your product”. Try telling that story around the water cooler.

I wrote this column for Search Engine Land’s Industrial Strength column. The idea came out of a presentation I did at SES. I was talking about the way we think about the value of an SEO program at Yahoo! – and why it’s necessary for others to take a stab at it before they ask their own IT or Web Engineering group to implement SEO recommendations. So, here we go….

Making The Case for SEO

So what’s the ROI on this SEO program you’re recommending? What will it cost? How much revenue upside is there? How does this stack up against the fourteen other projects we’re working on right now?

Are these questions familiar to you? They are to me. I hear them in my sleep. I have had to answer each of them at some point (and most more than once) when making SEO recommendations both in-house and to clients. As SEO continues to grow as a viable marketing channel, you need to be prepared to answer these questions as well. Why?

Every SEO effort requires changes to your company’s website. And every time you change a website, it not only costs money, but also requires valuable engineering resources that could be doing something else. So, if you’re considering recommending SEO at your company, it’s important to think about the impact your SEO efforts might have for the business in real dollar terms. This is driven by the fact that product and engineering managers need to prioritize engineering resources to be sure that, at any given time, they are being used to drive the maximum value to the company.

The larger the company, the bigger an issue this becomes. In multiproduct companies, the scope of this issue is greatly expanded. In large, multiproduct companies like Yahoo!, the scale and scope can be daunting. Nothing gets put on a product roadmap without a business case, and the more compelling a business case, the higher the priority on the product roadmap. The good news is that the same practices that allow us to succeed in large, multiproduct companies, also apply to small and mid-sized companies just as readily.

Value your traffic

It’s important to have a notion of what a click from a search engine user might have on your website. At Yahoo!, we use the notion of Lifetime Value to help us value web activity across a large number of web properties with vastly different business models. In principle, this does not have to be a complex exercise, but it can get as complicated as you let it. The idea is to quatify the revenue generated by a user over the lifetime of that user, and to discount that revenue stream back to the present. Use whatever data you have and then refine the model as you learn more. Where you don’t have data, use your best estimate, but make sure to note your assumptions and challenge them often.

Mind the Gap

Once you know what a click is worth, you’ll want to quantify potential results from SEO. A few different approaches, have worked for us. At one end of the spectrum is where we build a predictive ‘clickspace’ model, using a targeted keyword list combined with some competitive ranking analysis and search volume data. The resulting number-crunching tells us how many clicks separate us from our various competitors at any given time, showing us where our competitive ‘gaps’ are. We use these gaps to measure the opportunity in clicks, and then dollars, that an SEO program could help us capture. This is very expensive exercise, and not everyone is set up to do this. At the other end of the spectrum is what we call the ‘30% rule’. In the absence of any research, data, or discipline, simply project your existing traffic, assuming you can measure it, increasing at a rate of 30% per year, plug that into your valuation model, and off you go (I recommend using this as a last resort). Of course, there are many reasonable choices between these two models.

Proritize

At some point, any proposed SEO work boils down to a list of projects that need to be executed. Be prepared to prioritize them. All of them. The reason is that they won’t all get done at once, or sometimes at all. Keep in mind that the degree to which you can prioritize will likely affect your success. In the end, managers are not likely to remember which projects were implemented and which were not, so prioritizing gives you the best chance of SEO success.

Be (a little) Conservative

Assume you will be accountable to the business case you build. Try to balance your need to get your project prioritized with the likelihood of actually achieving significant results. This will help you build credibility, as nobody likes to be sold a ‘bill of goods’. I can vividly remember one SEO plan that I projected very aggressively so it would get done. It did, and when the implementation yielded less than stellar results, not only did I get an earful from the engineering community, but it took more than a year to restore my credibility with them.

Reward

Every win in SEO is significant. Be sure to latch onto any successes and celebrate them. Whatever you do, don’t take any credit! Instead, when any SEO work is done, praise the product and engineering resources responsible for implementing changes. Take exerpts from emails where you see positive results recognized, and circulate them around the organization. Hang onto these! They will serve as currency for your next SEO business case, which you will no doubt need shortly.

I’m going to use this site to post my columns and other content that I think would be of general interest. In the course of my work I generate lots of information on paid and organic search marketing in a big-company environment, some of which is actually useful to search marketers in general, so my goal is to take that stuff and post it here. I am currently working at Yahoo! and writing a column with a few like-minded individuals for Search Engine Land, aptly named “Industrial Strength”. You can find it here. More to come….