Search Engine Rankings

SEO used to be like playing a video game. You could see your page rising in the ranks on Google. SEOs would brag about how fast they could get to #1. Getting to #1 was even a reasonable goal in those days.

All that has changed.

Now search engines use personalized search to show people things that are most likely to be useful for them. This is better for searchers. However, it means that everyone sees different things. Your #1 is not always my #1.

Julianne, Rosie, and I fired up our computers and ran some searches to show you. We all live in the same town, with the same zip code, but we didn’t always get the same results.

Here, for example, is what I see when I search, signed in to Google, for “bakery”:

The first choice is the bakery nearest my home, where I currently am. They then offer me a New York bakery, a Wikipedia definition (seriously?), and then a fairly random collection of bakeries in my vicinity. Rosie and Julianne see the same thing.

Here’s what I see if I look for “bakery” while signed in at Bing, a search engine which I don’t actually use:

Bing clearly has no idea what I want here. I get a couple of definitions of “bakery” and sites with the word “bakery” in their addresses. Since I don’t use Bing often enough for Bing to get to know me, there’s no personalization going on.

Wikipedia is the only site offered by both Bing and Google when I’m signed in. Here are Bing and Google on Internet Explorer, a browser I rarely use — signed out, with the same location:

For a one-word search, Google offers just about the same thing to me, to Rosie, and to Julianne, whether we’re signed in or out. Bing’s results are different from Google’s, and quite different signed in from signed out.

In real life, we’re unlikely to search for a bakery in our town by using the single word “bakery,” though people try this all the time to see their business website’s rankings. Let’s try a more natural search.

When I search for “bakery fayetteville ar,” while signed in at Google on my usual browser, I see this:

I get some local results, now that they’re sure I want to find a bakery.

Same search, signed in at Bing:

Bing has access to my Facebook information, though frankly it doesn’t seem to be getting much value from it. The list is much the same as the previous options Bing has offered, but different from Google’s.

Julianne’s list is once again very similar to mine:

Rosie’s is different, though:

Rosie doesn’t get any local results, apart from #1 Rick’s Bakery, and she is shown an article from a Little Rock paper.

Who cares? Bliss Cupcake Cafe should, for one. So should Stone Mill.

Out of nine trials, Rick’s Bakery was #1 in eight.
Bliss Cupcakes was #2 in five of nine trials — and not even above the fold in the other four.
Stone Mill was #2 twice and #3 twice, and showed up below the local results twice, but was invisible in the other two trials.

No other local bakery showed consistently in most of the trials, though many more showed up in one or two.

These results show some truths about search rankings:

Personalized search doesn’t change rankings. Rick’s Bakery doesn’t show up as #1 for every possible search, but it shows up ahead of most other bakeries most of the time. Rick’s does not, as it happens, have a well optimized up-to-date website, but they are the most established local bakery and they deserve their ranking. Bliss and Stone Mill both have much better websites — and they’re beating the pants off older, more established bakeries online. It’s pretty clear that Google has ranked Bliss’s website ahead of Stone Mill’s, and I could certainly make some guesses why.

Personalized search may not matter so much when you’re #1, but it matters a lot if you’re not. All three of the top placing bakeries are active in social media and have websites (some local bakeries don’t). The Little Bread Company, which only showed up in a couple of our trials, may actually be a better bakery than some that do better online, but they have an unoptimized DIY website. Rick’s can get away with an outdated website, but the second-tier competitors can’t. Little Bread Company is good on social media and they probably show up well on some people’s computers — a better website could push them up for the rest of us. Certainly, Stone Mill’s position just a little behind Bliss could be affected by a website update.

Personalized search is influenced by so many factors that you can’t control it. Julianne and I get similar results from Google on local business searches (or at least on this one) no matter what, but Rosie’s are completely different. The overall ranking of the top three — Rick’s, Bliss, Stone Mill — is the same, but what she sees on the search engines results page is different enough that she might be influenced to click through quite differently. Her computer has spent more time in Little Rock than Julianne’s or mine, so Google may not be sure that she wants to go to a bakery right now — she might be doing abstract research. Who knows? Since the behavior of the individuals, their friends, and so forth affect the results, you can’t influence them nearly as much as you can influence the quality of your website.