Archive | rank weighted search results RSS feed for this section

eBay questions value of advertising on Google

14 Mar

eBay questions value of advertising on Google

A report by eBay finds that paying for AdWords does not boost its sales and that users mostly click through to its site via the organic results. “Removal of these [Google] advertisements simply raised the prominence of the eBay natural search result. Shutting paid search advertisements closed one (costly) path to a firm’s website but diverted traffic to the next easiest path (natural search), which is free to the advertiser.”

As Dr Philip Alford, director of Bournemouth University’s School of Tourism Digital Hub noted, the size of the brand makes a big difference to the effectiveness of paid searches. For SMEs, the take away is that organic optimisation on all your keywords, branded and non-, will bring better sales and brand impact ROI than will keyword buys. And if you’re going to buy keywords anyway, make sure your landing pages are optimised because Google takes that into account in its Quality ranking — which partly determines both your ad positioning and what you pay for it.

87% of all clicks go to top five SERPs

11 Mar

The top five SERPs garner 87% of all click-throughs

A recent search results click-through study from Compete.com finds that 53% of organic search clicks go to the first result and 87% of all clicks go to the top five.

That’s a key benchmark for online commerce — if you’re not in the top 5 results, you’ll miss 87% of potential click-throughs to your site. But if your area of responsibility is for something other than commerce, you might want to take a broader view of the search results positioning.

The entire first page of results has value for brand visibility and awareness because all ten results, not just the first five, are presented at the same time on one page. They create an opportunity for users to see your brand name (or your competitors’), even if they don’t click through to a site. So the brand needs to be prominent in the top 10 results, even if it’s not in the top 5.

The first two pages of search results (the entire top 20) should have the attention of executives responsible for PR and reputation management, investor relations, human resources and even legal counsel because what’s on the second or third pages one day (negative articles, regulatory filings, lawsuits, postings from disgruntled customers or ex-employees — can be bumped up onto the first page the next day. Bad news travels faster online than good news; negative results tend to climb up the rankings faster than positive ones. The results on the third, fourth and fifth pages are like the canary in the mineshaft — and early warning system — which is why proactive companies try to claim as many of the top 50 results as possible with their own or at least friendly content.

Recommendation: for online commerce, focus on the Top 5 SERPS; for branding, the Top 10; for PR, RM, IR, HR and legal, at minimum the Top 20.

Predictive Power of SERPs

11 Dec

The Predictive Power of Search Engine Results Pages (SERPs)

The top-ranked brands in those “Best/Most Valuable Brand” surveys almost without exception claim a significantly higher percentage of the Top 20 search results for queries for their own names than lesser ranked brands do. A reasonable explanation is that better-managed brands extend best practices throughout their entire operations, including online. And it’s a self reinforcing cycle: the more visible they are in the Top 20 search results (the only ones that really count) with positive, company endorsed content, the more traffic they get, the more customers they win, the more attention they receive from the media, analysts, investors, opinion researchers, and partners.

When measured over time, a brand’s performance in the Top 20 results can not only indicate whether it is maximising its online opportunity but can also anticipate increasing or declining interest from consumers, journalists and other commentators, which in turn is valuable knowledge for the investment community and acquiring companies. Of course, search results performance is only one metric of overall online performance. There are many others, but they key point about SERP performance is that the search engines tend to factor in those other metrics (broadly, metrics of strength and popularity) when ranking brands in their results. Tip: don’t just score the brand name; score the brand name with words like, e.g. complaints, scam, ripoff, unfair, layoffs, and other terms that suggest customer or employee dissatisfaction. And, compare several brands with similar market characteristics or financial fundamentals side by side. Look for company owned webpages, social profiles and apps, for desktop sites and mobile sites.

A brand’s weighted SERP score is thus more than a proxy for its historic overall online performance: it’s actually a predictive indicator of improving or deteriorating prospects.

Limitations of SERP and KW position checker tools

2 Nov

Limitations of SERP and KW position checker tools

SERP and KW position checker tools are designed for tracking where individual website pages appear in the search results for specific keywords. There are several small issues and two big ones associated with how they go about this.

  • they check for the position of a site or page for a KW query, so you have to enter a URL AND KW (they can’t check for just a KW, or just a site). That’s inflexible.
  • on many of them, you have to do your checks one at a time, on a new interface for each. That’s time consuming.
  • only a few give you downloadable files (you then have to merge all those files, when you can get them). Most return your results on html or dynamic pages (copy and paste). That’s unwieldy.
  • they vary in the number of maximum positions considered. Some show 100, some more, some less; and they are often out of date. That’s unreliable.

Turning to the bigger issues, most crucially, SERP and KW position checker tools don’t tell you how they generate their results, or how fresh they are.

Data pulled from an API is only as good as the API itself. Our experiments with an API were not encouraging (which was why we dumped it a while ago): It gave lower lower counts than browser based counts — it showed on average 30% and sometimes 90% fewer pages indexed than a browser search; and it was less accurate than browser search results — the API SERPs failed to match browser SERPs by on average 10% and sometimes as much as 90%. You cannot statistically compensate for errors of this magnitude in both quantity and quality, even with sophisticated mathematical optimization routines.

Another common method, scraping, has problems of its own: The rotating IP proxies necessary to carry it out will skew the results because each successive query will be seen to originate from a different location than the last one — a different city, sometimes a different continent. That’s useless if you’re running more than one KW for a brand.

Finally, SERP and KW position checkers show the numeric position of a page returned for a specific query in the search results. There are two limitations to this:

  • Consumers search for brand names and/or keywords (which they enter into the search bar), they do not enter website addresses into the search bar (if they know the URL, they enter it straight into the URL bar, bypassing the search function).
  • Knowing the numeric position of a particular page is tactically useful but is not strategically insightful. It doesn’t give you a snapshot of who claims the Top 20 results for your brand name and keywords, nor a measure of your overall online strength.

With RankTank, you can check and track your SERPs while aslo scoring and ranking them.

Corporate Responsibility Magazine’s 100 Best Corporate Citizens cross-compared for brand name + corporate citizen against score for brand name only

18 Oct

RankTank took the 23 brands from Interbrand’s 2012 Best Brands report which are also on Corporate Responsibility Magazine’s 100 Best Corporate Citizens 2012 list and cross-compared their score for brand name + corporate citizen against their score for their brand name only.

Of the 23 brands on both Interbrand’s and Corporate Responsibility Magazine’s lists:

5 have a better search results score for their corporate name + corporate citizen than for their name corporate alone and are overperforming for this topic (Chart 1).

  • Disney scores only 38 for its corporate name alone, but 78 for its name + corporate citizen — a positive gap of 40.
  • GE scores only 51 for its corporate name alone, but 65 for its name + corporate citizen — a positive gap of 14.

Disney and GE are ranked #13 and #6 respectively by Interbrand (Brand Value), and #2 and #19 by RankTank (Online Strength). They are among the best online performers for “corporate citizen”.

18 have a worse online strength score for their corporate name + corporate citizen than for their corporate name alone and are underperforming for this topic (Chart 1).

  • Nike scores 76 for its corporate name alone but only 10 for its name + corporate citizen — a negative gap of 66
  • Colgate scores 82 for its corporate name alone but only 22 for its name + corporate citizen — a negative gap of 60.

Nike and Colgate are ranked #26 and #47 respectively by Interbrand (Brand Value), and #30 and #97 by RankTank (Online Strength). They are among the worst online performers for “corporate citizen”.

See who’s overperforming and underperforming against their corporate citizenship investmwent at http://www.docstoc.com/docs/133642764/Best-Corporate-Citizens-ranked-by-Online-Performance

Companies in Dow Jones Global Sustainability Index Ranked for Online Performance for “Sustainability”

18 Oct

RankTank re-ranked the Top and Bottom 20 companies in CSRHub’s Dow Jones Global Sustainability Index (390 companies covered), comparing their ranking awarded by Dow Jones to their online strength using the rank-weighted search results claimed by each company as a proxy for performance on the topic of “sustainability”.

Out of 2,000 search results available to them for a search for their corporate names and “sustainability”, the 20 companies ranked highest out of 390 by Dow Jones for sustainability:

  • collectively own 920 results
  • give up 781 positions to “Others”
  • have an average ECR score of 46 (low-mid range of online performance).

Out of 2,000 search results available to them for a search for their corporate names and “sustainability”, the 20 companies ranked lowest out of 390 by Dow Jones for sustainability:

  • collectively own 650 results
  • give up 908 positions to “Others”
  • have an average ECR score of 35 (low end of online performance).

Find out which companies are getting the the most visible return on investment for their sustainability spending at http://www.docstoc.com/docs/133642796/Global-Sustainability-Index-ranked-by-search-engine-performance.

Scoring Brand Online Strength by Rank Weighted Search Results

10 Oct

Scoring Brand Online Strength by Rank Weighted Search Results

I first started weighting search results positions in October 2009 with a system that weighted the Top 20 results pages 20 – 1 (#1 result = 20, # 2 = 19 and so on down to #20 result = 1), for a total of 210 “points.”

RankTank's weighting system 2009 - 2012

This assigned 74% of the value of the Top 20 results (155 points out of 210) to the first 10, and 26%, (or 55 points) to the second ten, in a straight line progression.

RankTank's weighting system 2009 - 2012 weights

RankTank’s weighting system 2009 – 2012 weights

In actuality, results page values do not fall in a straight line progression but in a stepped one so in 2012 I modified the weights to the values shown in the application, which are:

RankTank's improved rank weighting system 2012

RankTank’s improved rank weighting system 2012

The new system assigns 66% of the value of the Top 20 results to the first 10, in a stepped progression that reflects how users interact with the results page in terms of a) scanning the list and b) actually clicking on results. By upping the value of the second set of ten, it also recognises their potential to move up into the first set of ten.

RankTank's improved rank weighting system 2012 position weights

RankTank’s improved rank weighting system 2012 position weights

Our business is scoring a brand’s online strength, so it’s important to score the degree to which a brand is protecting its top 10 position by blocking out the next set of ten with pages it also claims: Today’s #19 result can be tomorrow’s #9.

Scoring Brand Online Strength by Simple Count/Share

10 Oct

Scoring Brand Online Strength by Simple Search Results Ownership Count/Share

I first discovered the relationship between search engine results page (SERP) performance and company performance in 2006. The first “top brand” source I used was the Interbrand/BusinessWeek Top 100 companies survey published August 7, 2006 (http://www.businessweek.com/magazine/toc/06_32/B399606globalbrands.htm). I ran a Google search on each company (I queried the exact name, without keyword modifiers or qualifiers) and classified the Top 20 SERPs according to who claimed the results of a query for its name: the enterprise (EGM), mainstream media (MSM), consumer-generated media (CGM), or others (OGM).

In 2006, I did not rank-weight the search positions; I scored using a simple count, treating all Top 20 positions as of equal value. I changed to rank weights in 2009.

Best Brands of 2012 continue to dominate online as well as offline

3 Oct

The Best Brands of 2012 continue to dominate online as well as offline.

On 2nd October, 2012, RankTank compared the online performance of 100 leading brands in Interbrand’s 2012 Best Global Brands report with their position in the brand league table.  As is customary, we use https://ranktank.wordpress.com/2012/09/30/weighted-search-scores-beat-simple-counts/ rank weighted search positions as the measure of online performance, specifically, the percentage of the Top 20 results each brand claims for a query for its own name.

Top level findings:

In 2012, the 100 Best Brands collectively claim 65.6% of the total available value of the Top 20 search results positions. This is consistent with other analyses we’ve undertaken since 2006, which show that on average the top brands claim 66% of their total available search engine value versus 30% for laggard brands.

RankTank rank weighted analysis of Interbrand 2012 Best Brands

RankTank’s rank weighted analysis of Interbrand 2012 Best Brands

The full report is at http://www.docstoc.com/docs/133640194/Measuring-Online-Performance-Report. It contains surprising conclusions on social media and a comparison between the 100 Best Brands performances in 2012 and 2006.

All brands were scored on Google.com Tuesday 2nd October between 1:30pm and 7:30pm GMT +1. We queried the exact brand name as in the report, without disambiguating or modifying terms.

Source: http://www.interbrand.com/en/best-global-brands/2012/Best-Global-Brands-2012-Brand-View.aspx

Weighted Search Scores beat Simple Counts

30 Sep

For organic search results, a weighted score is better than a simple count

Simply counting the number of pages your brand claims (owns or controls) in the top results does not provide a sufficient measure of your brand’s actual online performance. Why? Because not all results positions are equal [link to Don’t just Google your brand, Rank it]. Here’s a simple example. Say your brand owns or controls the #s 3, 6, 7, 8 and 10 results. A simple count would be 5 out of 10. But that’s misleading because it ignores the value of higher ranked results over lower ones. (The #1 result in Google is there for a reason; so is the #5,789,645.) RankTank assigns weights to each result position as follows (below we show just the Top 10 results as an example):

table of search results positions versus weights

The value of a search result depends on its position in the rankings

Now calculate your weighted score in RankTank.

RankTank calculation of rank weighted score

RankTank calculation of rank weighted score

The results you claim (Brand A’s ECR, the weighted sum of “e”) are worth a total score of 29.

The results your competitor claims (Brand B’s ECR) are worth a total score of 37.

[For an overview of ECR, etc, see http://ranktank.eu/How-To-Use-RankTank.htm. For a detailed explanation, see http://www.slideshare.net/RankTank/rank-tank-tutorial-14524415.]

So on a simple count basis, you’d claim half the results for your brand name, apparently a fifty percent score, but on a rank-weighted basis, which reflects what users actually do with search results, you’re underperforming your competitor.

This example is based on the Top 10 results, not the entire 20. As it shows, out of a possible score of 66, your brand scores 29 and has left 37 on the table for dissatisfied customers, disgruntled ex-employees, brandjackers, and of course competitors. The differentials are even larger when ranking the Top 20, which has a total possible score of 100.

Now repeat the above — for all the keywords important to your company’s operations (sales, PR, IR, HR, legal, advertising, CRM, partners, etc) — in the top 20 results across all search engines, and you’ll get a sense of the Share of Online Voice you’re losing out on if you’re simply counting the number of results instead of using RankTank weighted scores.