Myth Busting the ROI on SEM

I’ve seen several different citations on why SEM has a better ROI then SEO.   Here are a few of them:

• SEM is faster, brings traffic to your website quickly, and increases visibility immediate
• SEM: there is more control over SEM then SEO.  With SEM you can optimize the landing page tailored to the SEM experience
• People who come from SEM are more likely to buy, People coming through PPC are more like to spend more on that site and the NY times article referenced
• The Average value of a visitor from SEM is more than the average value of a visitor from SEO

I’m also sure you every go to a conference and the topic of SEM comes up you will hear similar stories.  What has to be looked at is what is being presented.  When it comes to numbers things can get very tricky very fast.  So let’s break each of these myths down, and in some cases finish off the missing part of the math.

SEM is faster, brings traffic to your website quickly, and increases visibility immediately

The search spiders today are pretty quick.  If you have a website you can get indexed almost instantly by submitting an XML sitemap to it.  You do have a sitemap right?   So lets assume then that faster here means it’s faster for you to implement. OK that may be true.  But now you need to manage your SEM campaign.  SEM should never be a “set it and forget it” initiative.  If you are committing to SEM, you should be reviewing your account minimum once a month and more preferably ever week or every other week.  Unless that is, you like to throw money away.  While you may have the campaign up and running quickly, plan for more nurturing of it over the long run.  Your site map, if you automate and create the proper rules you can “set it and forget it”, but you will still have to look at your other SEO elements, so in my mind this is a wash in the worst case scenario.

SEM brings traffic to your website quickly, and increases visibility immediately – yup it sure does.  So can getting on the front page of Digg or other large news aggregator sites.  Does this spike in traffic mean its good traffic?  No.  Not all traffic is created equally.  You need to make sure you are not bidding on the wrong terms, that the page you are driving people to is converting at its optimum rate and that anyone who is researching you can come back and find you.

To accomplish this you better be prepared to spend, and then spend some more.  You pay for every click.  That means you are paying to learn from your mistakes.  If you see lots of traffic but no sales you need to figure out if the problem is with the words you are bidding on or your website.  This can be an expensive lesson to learn for a new company or a company just getting into search, also if someone is comparing several companies before deciding on who to buy from, you better budget to be in market all day every day, so if they pick you as the one to buy from they can find your site again in the SEM results.  If you plan on running SEM as a new company expect your campaigns to simply be awareness campaigns. Don’t plan on making lots of sales, see it more as letting people know you are there.  Maybe you’ll make a couple sales maybe not.

If you are an established company this changes a bit, but then you should be doing OK on your SEO front.  You should have had time to develop some rankings in the SEO listings.  Maybe you are not ranking on some non-branded terms you think you should, and you are known enough in the space that people will recognize your brand.  Then yes use SEM to fill this hole, but start building an SEO plan to fix this in the long run.

SEM: there is more control over SEM then SEO.  With SEM you can optimize the landing page tailored to the SEM experience

This is just stupid.  I’m sorry but if you haven’t reviewed what your top SEO entry points are and looked at your bounce rate, then there’s a problem.  Pages that rank well can still be optimized.  Further you are likely ranking well in SEO BECAUSE your page does a good job of providing what a customer is looking for.  I think this myth is actually the fault of the SEOs of this world.  There is so much focus on optimizing a page for a search engine, or a machine, they forget to look at the human factor.  A page can be optimized for the engine, and rank well, and you can actually change things around on the page to try and improve the conversion even more.   You can use JavaScript to A/B test small elements that won’t get you penalized by the search engines right away, and incorporate the well working elements into the page.  My message for SEOs stop being paranoid of changing things because you have a top ranking, change and change again until you create a page with improved conversion rates.

Also all the elements of an SEM ad are present in the SEO results and manipulatable as you can see here:

Click to Enlarge

The description in Google is  influenced by the meta description tag.  The Title tag influences the title.  Just like in SEM you can tailor both of these.  If anything this is where SEM can help you out.  Test several copies and see which produces the most clicks, then change your title and description to that SEM variation, and see if it improves your CTR in SEO (you can check that out in webmaster tools)

If the page slips in the rankings because of a change you made, go and unchanged that page.  Finally just because you are number one today doesn’t mean you will be tomorrow, everyone under you is looking to move up in the rankings, they are changing and tweaking their pages to get a leap on you.  If you sit around long enough someone will eventually pass you in the rankings.  Long story short for SEO don’t be stagnant.  Tweak away!

People who come from SEM are more likely to buy; People coming through PPC are more like to spend more on that site

This isn’t even like comparing apples to oranges.  Here’s the problem.  If you are running a half decent SEM campaign you are likely trying to sell the products that produce the highest profits or the products with the bigger price tags.  SEO on the other hand is not discriminating.  It will cover your lower margin products, and lower priced products as well as your more expensive and better margin products.  SEO fills the gaps SEM is leaving, it is going to suffer the cost because of this and see lower average order values.  A better metric would be how many orders SEO drives and how many orders does SEM drive.  Don’t put a value on it, just look at shear volume, most likely you will see only 10%-25% of the total number coming from SEM.  Now take each order and convert it to the profit margin, and add up the entire SEO profit margin total and the entire SEM profit margin total.  Which has a higher TOTAL profit margin?  SEO may and I would suspect will have the higher margin because it encapsulates many more orders which are also referred to as the long tail of search.  Oddly enough SEM folks preach about the long tail for bidding on terms, but then want to forget about it with a metric like this.  Amazon makes more money off the obscure books they sell then all the best sellers combined.  So make profit off a couple big orders or lots of smaller ones, figure out what works for your business.

As for the time on site, that’s a number I could care less about.   It tells you nothing, it could be a lot of people sitting idly on the site.  Some people may say this is an idea of engagement; this is false.  You should measure engagement with interactions on a page.  Capture scrolls on the bar to see if people are reading to the bottom of your page or clicking on content, that’s a better indicator.

The average value of a visitor from SEM is more than the average value of a visitor from SEO.

So again, this is a very miss leading metric. The average value of a website visitor should be higher with SEM.  I mean you are paying for that click, you had better be driving customers to buy something on the words you purchase.  When I look at this metric though I think the same as time spent on site.  So what?

Most sites should expect to see 75%-90% of search traffic come from SEO and 10%-25% come from SEM.  So let’s assume 5000 visitors come from search total and I’ll even give SEM the better number of 25% traffic here (3750 from SEO and 1250 from SEM) we calculate the value of each by multiplying the value of SEM ($2.38) by the amount of traffic (1250) for a total of $2975.  Let’s do the same for SEO now 3750 visits multiplied by $1.35 (average value of a customer) for a total $5062.50.

Let me break that down.  SEM will only ever drive $2975 while SEO will drive $5062.50 in revenue.   A well indexed, and well optimized site will ALWAYS drive more revenue than SEM.

Factor in on top of that traffic from SEO is not capped; traffic from SEM is capped based on the budget you have to throw at it.  A well developed SEO strategy will always trump an SEM strategy.  Also you better not be paying more than $2.38 per click from SEM.  If you are then you will certainly be running into a negative ROI, and that is assuming the value per customer is based on profit margin, if it is based on the gross value, you have real issues, as your net margin will be even lower bringing that value down further.  The only true value I see PPC having is in testing.  So who wants to insist that SEM has more value than SEO?

There is also point and counter point to all this from some industry experts at: http://blog.searchenginewatch.com/091217-153103 personally I think the SEO folks did a bad job of defending SEO.