Do Search Engines (Google) “Harm Minority Owned Businesses”?

Yesterday WatchdogWatch.org, which monitors NGOs (“watchdogs”) found and blogged about a filing with the FCC in the larger context of its proceedings on net neutrality. The filing is by an organization representing minority owned media and business. The issue raised is whether the concept of “net neutrality” should be expanded to include search engines and appearance in search results:

The Minority Media and Telecommunications Council, representing 16 national organizations, [� ] raises the provocative issue of whether net neutrality requirements should be applied to search engines.

The MMTC report (.pdf) says the following about alleged search engine “bias” against minority owned businesses (and especially small business):

Google, recently accounted for over 71% of the entire U.S. search market,109 giving Google an unprecedented ability to influence this market. Controlling the results of such a substantial share of the Internet search market gives Google an ability to influence what websites are accessed and thus which web based businesses are able to generate the traffic necessary to succeed. The National Organizations are concerned about reports that Google is “heavily biased towards established brands” and erects “significant barrier[s] to new entrants and inevitably suppress[es] innovation.” This trend could have a particularly harmful impact on minority-owned businesses, many of which are attempting to gain a foothold on the web while lacking access to capital. Search engine practices that assign visibility to businesses based on wealth rather than merit would impose a classic cycle of invisibility to minority enterprises: without access to capital they cannot secure visibility; but without visibility they cannot secure access to capital.(emphasis added.)

There are a number of assumptions in this set of claims that may either simply be false or merit actual investigation before any meaningful discussion can occur. For example there’s a leap of logic that argues SERP visibility equals access to capital. Hockey stick growth for an internet startup, which may depend on SEO in part, could mean access to capital/funding. But access to funding or loans at the small business level is going to be independent, for example, of Google or other search engine visibility.

In addition, the notion that Google or search engines more generally are “heavily biased toward established brands” would be news to News Corp’s Rupert Murdoch, who might argue that they should be. Indeed, AP and others have made similar arguments. See for example:

Big Brand Media Wants A Google Bailout Sorry, Tom Curley: Don�t Expect A Google Ranking Boost For The AP

I spoke to Bruce Levinson at some length about all this. Levinson is with the Center for Regulatory Effectiveness, which publishes WatchdogWatch and wrote the post above. He recognizes the absence of facts behind the claims in the filing. His view is that there should be an investigation of whether the injuries (or hypothetical injuries and bias) described in the report actually exist. After speaking with him I would agree that some sort of study might be worthwhile.

I told him, however, that Google has no way of knowing which businesses are minority owned (large or small) and that its algorithm doesn’t and in fact can’t consider this factor. In the context of its local algorthim, Google’s increasing bias is in favor of local, small businesses and against larger, wealthier entities (e.g., aggregators, directories). But let’s assume for a moment that some sort of empirical study found that minority owned businesses were in fact disproportionately “less visible” in Google search results.

What would the “remedy” be? We quickly get into analogies to affirmative action, which would be problematic for all involved. Would search engines need to rectify imbalances by offering privileged status to certain types of results. This would bring a massive outcry.

To reveal my own bias, I support affirmative action but feel that any program that sought to favor certain kinds of businesses or search results would fail on multiple fronts. The problem, if there is one, is better addressed in the “real world” in terms of monitory business programs and access to capital. But it may be worthwhile to determine whether search engines and SERPs do in fact show “bias” or have a “disparate (adverse) impact” on minority owned businesses.

Postscript: I received an email after posting this story that makes the valid point that entities with more resources for website design and promotion are going to rank better than those that have fewer resources. This is a very legitimate point. The argument is: this “bias” toward “well designed” sites ultimately disadvantages smaller companies vs. larger ones. That may well be.

However it’s another question as to whether, among small companies, there’s a specifically “disparate impact” or adverse effect on minority owned businesses.

No comments:

Post a Comment