Note: this is a post from 2007 – that was orphaned from it’s original context; and is being republished here.
How can we determine if Search Engine Optimization is a good value?
The answer is not simple, but dependant on many factors. The first factor we’ll discuss is the value of a new customer or transaction. With eCommerce, this is much simpler, by the virtue of the fact that we can track a direct correlation between the cost of acquiring a customer, and the amount that is spent on a transaction. Even there, however, we must consider the lifetime value of a customer. If 20% of our first-time buyers return within the next two years to make an additional purchase, we must factor that into our calculations.
For every dollar coming into an organization, there is a cost associated with acquiring that dollar beyond production costs. We include here marketing costs, advertising costs, and sales commissions.
It is important for a business to have a clear understanding of what costs they are willing to bear for each dollar of income. It is likely, in fact, that different product or service offerings within an organization are likely to have different costs.
Acme Corporation sells Widgetos, a wonderful new widget product. The cost of manufacturing, packaging, and distribution for each Widgeto is $1.00. If the business plan calls for a 10% profit off of each Widgeto sold, and the competitor’s product is sold at $2.00, perhaps we’ll want our product to be sold at $1.90 (unless we’re in the luxury goods industry – in which case we might want to price it higher!).
$1.90 – 10% profit ($0.19 per unit) = $1.71
Minus cost of product, $1.00, = $0.71
Our Chief of Marketing would be willing to budget $0.71 for the acquisition of each sell. Now, it just so happens that this product is perfect for selling on the web. With our Pay-per-click (PPC) campaign, its pretty simple to plan it out, where if one out of seven visitors through a PPC advertisement make a purchase, and each of those clicks costs $0.10, we’re ahead of the game. If it takes 10 visitors, we’re losing – or if the cost per click goes above ten cents, and still only one out of seven visitors makes a purchase, we’re not gaining. But at least we can see pretty clearly where we need to make improvements; by either improving the conversion rate (how many visitors make a purchase) or by lowering the cost per click. That is the task of the PPC Manager – and there are many tools in the toolkit for doing that.
With SEO, the exercise becomes a bit more complex. First, with SEO, there is a longer life cycle for appearing higher in a search engine. This life cycle can vary with different key phrases – and is usually related to the competitiveness of a key phrase. On less competitive key phrases, it is not uncommon for a website to continue to rank highly for several years. With a more competitive key phrase, it may require more ongoing maintenance, but it is also possible to remain high in the rankings. But what we’re trying to determine here is the lifetime value of the base set of SEO activities.
Another wrench-in-the-works with determining the ROI of SEO is that SEO usually produces a lot of visitors with related key phrases. I might have optimized content for “Manhattan Honda Dealer”, and find that I’m getting visitors who have searched for “Honda dealer in Manhattan” or “Honda civic new Manhattan”. There is a cumulative effect with SEO that spans both time and key phrases. And added to that is the notion that in a sound SEO initiative, more and more targeted key phrases will be added over time, which again has a cumulative positive influence on the rankings of those initial key phrases.
What this means in determining ROI is that we must make some broad assumptions, basing them on averages and likelihood. But by understanding each of the factors, the savvy marketer can test the data against the assumptions, and help steer the strategic course.
Cost of SEO Project
Let’s start by looking at the initial cost of an SEO project. Lets say that a basic project has a setup cost of $2,200, with six months of ongoing activities at $500 per month. Over the course of this project, there has been a cost of $5,200.
We show an increase over 3 years as this is a pattern that we’ve seen. Some key phrases will deliver more visitors quickly or evenly over a long period of time.
|Year 1||Year 2||Year 3|
Over the course of 3 years, we’ve received 1,200 visitors from natural search on those key phrases, or related key phrases. Our cost per each user has been:
$5,200/1200 = $4.33 per visitor. Now, if one out of ten visitors becomes a customer, the cost for each customer has been $43.30. Depending on what we were selling, this has been a positive return on investment or not. There are businesses that would gladly pay thousands of dollars for each new customer, and there are some that would pay less than $1.
Of course, for the sake of this discussion, I’m not getting into the “cost of capital”. Perhaps our business owner had to borrow that $5,200, and is paying interest on that loan over 3 years.
In our shop, we usually structure our SEO projects on 20 key phrases at a time. Our KEI research makes use of a series of tools, including Wordtracker, Trellian, and the tools provided by Google. The chart below shows a typical KEI report (showing only Google) for a series of key phrases for a PR firm:
|Key Phrase||Searches in last 90 days||# of
|relative market share profit||378||403||354.6||8460|
|research survey entrepreneur||128||116||141.2||2850|
|crisis communication for attorneys||29||25||33.64||630|
|the five minute presentation||195||1830||20.78||4350|
|email marketing consulting||328||9650||11.15||7320|
|crisis communication for lawyers||17||89||3.247||360|
|business promotions and marketing ideas||20||147||2.721||420|
|hazard communication plans||40||601||2.662||870|
|organizational functions of pr||32||517||1.981||690|
|emergency communication disaster plan||11||73||1.658||240|
|societal functions of public relations||45||1340||1.511||990|
|newsletter writing service||24||563||1.023||510|
|media integrated training services||6||38||0.947||120|
We need to consider a couple of different things here. We are looking for key phrases that we have a high likelihood of getting in the top search engine results. If we target key phrases where we’re unlikely to get in the top 10, our results won’t be very fruitful.
Our next assumption needs to be this: after considering how many searches there are for our chosen list of key phrases, how many clicks are we likely to get if we’re in the top 2,3, or 4th position? Or the first position?
I might make an assumption like this: My set of key phrases is searched for by 500 searches per month, or 6000 per year.
Out of that 6000 per year, I’d be looking to gain a tenth of that traffic, or 600 visitors per year, over a three year period. Remember, it is common for us to get a lot of traffic for similar key phrases too.
In some campaigns where we’re trafficking extremely “long-tail” key phrases, our traffic might be a fraction of that – but it might be for a business where the acquisition value of each customer is very high.
Another aspect of SEO ROI that is difficult to measure is in “Latent Conversions”. Research has been showing that there are several scenarios where SEO provides a boost to PPC and other marketing. For instance, it is not uncommon for buyers to search on a phrase, only to return and search for a specific branded term, clicking on the ad. We credit the PPC ad with the conversion as we’re not able to identify the earlier search. With some key phrases, it is not uncommon for us to see the natural result directly beneath the paid result. In these circumstances, the searcher is being presented with the customer’s website several times!
One of the main activities used in the SEO toolkit is the building of “back links”. These are links from other websites to our website, hopefully using the appropriate key phrases, and pointed to the appropriate page. While we’re watching the increase in traffic coming from the Search Engines, we might notice traffic also coming from those back-links.
To Michael Coppola from Path Interactive for his generous feedback