October 21, 2009

Jedi Metrics: How To Overcome Resistance To SEO Efforts

(Read in entirety at Search Engine Land. Published 2009)

Last time, I shared a set of natural search metrics that help frame your natural search business opportunity and win the resources needed to improve your SEO performance. In this article, I’ll show how to use the first three metrics to begin building a kick-ass business case that your executives will find hard to resist. Before we proceed, just one word of advice from Obi-Wan, young Master Luke: Only use these metrics to sell (either internally or externally) if you have ability to deliver on what you’re selling.

Define the market opportunity

First, what’s the total size of market you’re chasing in natural search? Yeah, I see your blank stare. This number has to be your mission; it must be top of mind in order to successfully convince executives to invest in SEO. Think of this first metric as a basic marketing feasibility study. The SEO costs and tactics are irrelevant at this stage. Tell me how many people are searching for what you sell every month. What keywords do they use—besides your brand name. This exercise is different from traditional “keyword research.” We’re looking for an estimate of market demand.

To estimate keyword market demand, first look at the phrases currently driving traffic. Export (from your site analytics) all the non-brand phrases that drove natural search traffic over the last completed month. Augment this list with all your PPC phrases that outperform your average site conversion. You can further augment the list by running a spider over your site to extract onpage phrases that may be worth exploring, but this is a more advanced step and is probably counter-productive at this stage. (Trust me: your business case will be compelling enough by the time we’re done. Resist the temptation to inflate it into unrealistic fantasy data that will only hurt your credibility later.)

Historically the tools needed to conduct our feasibility study have simply been unavailable, or we were forced to make keyword research tools do a job they weren’t meant for. Thankfully, Google made its AdWords Keyword Tool, with demand data, accessible for advertisers over a year ago. Hopefully Yahoo and Bing will follow. For now, simply feed your keyword list into the AdWords tool over the comparable time frame. You can query using broad or exact match, but I’d suggest exact match to calculate more precisely how much of that market you penetrated.

A few caveats here: First, Google’s numbers are imperfect, but better than the alternatives. We find that Google will fail to report useful data for many long-tail keywords that have small demand. This is an issue since long-tail keywords are the essence of what big natural search programs are made of (another tool limitation, for the time-being). Second, we suspect the demand numbers Google reports are overstated by a factor of between 20% – 40%. So you can discount accordingly to keep a clean, consistent Google-only calculation through your business case.

At the end of this process, you want a spreadsheet with the number of keyword phrases and the demand of those markets, with a total market size that you are competing in, across natural and paid search. You have defined your market opportunity when you can confidently tell your executives something like this: “Our addressable market spans 200,000 different keyword markets, and totals roughly 5,000,000 searches per month.”

Estimate click through rate (CTR)

The next question asked will be, “But how much of that are we getting right now?” To answer this, determine what percentage of those markets your current keyword traffic represents. This is an easy and powerful calculation. You know what non-brand traffic you received from Google during the same time period (from your site analytics). Simply divide the market opportunity by actual traffic received for each keyword to determine your CTR for each keyword. You should calculate CTR on an individual keyword-level basis. For bottom-line purposes, calculate a weighted average of the individual keyword CTRs.

You have defined your CTR when you can now tell your executives something like this: “Our addressable market spans 200,000 unique keyword markets and totals roughly 5,000,000 searches per month, of which we’re currently capturing 1%.”

(Pause the movie a second. Did you catch what just happened? With just two metrics you now have the power to compare the efficacy and potential of natural search against other forms of online advertising using familiar performance metrics—without resorting to ineffective faith-based arguments or resource-begging. It gets better. Unpause…)

For added power, insert a column in your spreadsheet showing where your site ranks for each phrase (SERP page is fine for now). As you might expect, you’ll find that phrases for which you’re ranked on Page 1 of Google (here I’ll be using shorthand like P1, P2, etc.) will have a higher CTR than phrases for which you are ranked deeper. Now compare CTR on P1-ranked phrases against P2-ranked phrases, etc. Our retail clients find Google P1 consistently achieves CTR of around 10%, while P2 achieves around half the CTR (<5%). If you see similar values, the business proposition is crystal clear: There are performance gains to be had by investing in phrases ranked deeper than P1. We all know this instinctively, but now the data can do the talking for you. Note: There are tools that report on SERP rankings automatically. Or you can parse referral strings yourself on all engine traffic to capture what page of the SERPS the searcher clicked on to land on your page.

If you let it, the data can say something even more powerful. Once you quantify the performance difference between your P1 and P2 (and deeper) phrases, you can assign and model a budget value to achieving the desired lift across the desired markets—along with some probability assumptions. Don’t be surprised if that budget is much more than you’re budgeting for all your SEO efforts right now. You’ll recall my premise: your SEO is underperforming now for this very reason.

Map out traffic acquisition cost

Now for the trifecta: Knowing your addressable market size and current penetration, your executives want to know two things: How much are you spending on that performance now, and how much does “more” cost. You’re in the game. The costs of advertising via paid search continue to increase. The cost of advertising through natural search is dirt cheap. Your executives want to increase profit by acquiring more traffic (and sales) at less cost. Natural search is an obvious way to increase traffic, decrease costs and increase profit, right?

You can easily answer the “cost of more” question by exporting the average CPC value reported by Google’s AdWords Keyword Tool above. Simply add two columns to your spreadsheet, one to reflect a growth assumption on your current natural search traffic (50%, 100%, whatever), the other to estimate what it would cost to acquire that traffic using AdWords. Total it up. You’ve just created a reality-based framework for valuing natural search that executives will appreciate – even though it may total more than what you’re currently budgeting to spend on SEO.

Your message to the CFO now looks like this: “Our addressable market spans 200,000 unique keyword markets and totals roughly 5,000,000 searches per month, of which we’re capturing 1% currently. The cost of acquiring that traffic through PPC totals $1,000,000 per month.”

More can be said on this point in future installments. But notice that you’re no longer taking Quixotic approaches towards SEO. Instead you’re making a business proposition based on markets with known performance and that have real costs associated. Your data is doing the talking for you, and your CFO’s heart is warming up to you. Powerful, the force in you is, yes…

Next time I’ll discuss the remaining metrics that help measure incremental gains in performance.

Submit a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Posted By


Search Engine Land