What Is Benchmarking?

The German business dictionary Gabler Wirtschaftslexikon defines benchmarking as the continuous comparison of products, services, processes and methods with several companies, in order to close the performance gap to the best in class systematically. The underlying idea is to find out what differences exist, why they exist and what improvements are possible.

In a 1993 interview with The Healthcare Forum Journal, Robert C. Camp said that formal benchmarking started at Xerox in late 1979 and early 1980: "We believe that we essentially invented it at Xerox." Xerox called benchmarking "the continuous process of measuring our products, services, and practices against our competition or leading-edge companies."

In the same interview, Camp distinguished four types. Internal benchmarking compares your own processes with similar ones elsewhere in the organisation. Competitive benchmarking measures you against others in your own industry. The third type compares with companies that are not competitors, for example in billing. The fourth, generic process benchmarking, asks for each process who does it best. APQC, a US organisation devoted to benchmarking, now lists other types, including performance, process, functional and strategic benchmarking.

Gabler describes five steps. First, choose the object to compare, such as a product, a method or a process. Next, choose the comparison partners and settle which similarities make the comparison valid. Third, gather data, from secondary sources or directly, for instance on site visits. Fourth, identify the performance gaps and their causes. Finally, set and carry out the improvements.

In marketing and SEO, what gets compared is usually metrics: search visibility, rankings for particular keywords, click-through rates, conversion rates or cost per customer. Tools such as the SISTRIX Visibility Index make the visibility of different domains comparable. SISTRIX describes it as a metric for a domain's visibility in Google's results, calculated from organic rankings in the top 100 for a fixed keyword set, weighted by search volume and position.

These values are a provider's estimates based on its own keyword set, not a measurement of traffic. They work for comparing domains with each other and over time, not as a substitute for your own Search Console data. Comparing a KPI or ROI with other companies only works if both are calculated the same way.

Without a comparison, no figure has a scale. A conversion rate of 2% is good or bad depending on what your own past, other parts of the company or comparable providers achieve. Benchmarking supplies that scale and shows where effort pays off most.

Benchmarking with competitors has a limit in competition law. The European Commission's 2023 Guidelines on horizontal co-operation agreements acknowledge that undertakings may become more efficient by benchmarking their performance against best practices in the industry (paragraph 425). For the purpose of benchmarking, however, an exchange of individualised data would generally not be indispensable, because aggregated information, such as an industry ranking, could generate the same efficiency gains with a lower risk of collusion (paragraph 426).

A second limit concerns publication. Advertising that cites results identifying a competitor is comparative advertising (Article 2(c) of Directive 2006/114/EC). Under Article 4, it is only permitted if several conditions are all met; among them, under point (c), it must objectively compare material, relevant, verifiable and representative features. A benchmark whose data and method you cannot disclose is therefore no basis for claims about other companies.

• Before comparing, settle what is being compared and which similarities make the comparison valid.
• Start with the internal comparison and your own earlier values, which you know in full.
• For external data, record the source, period, measurement method and sample.
• Label values from SEO tools as estimates and use Search Console for your own traffic.
• Exchange only aggregated or public data with competitors and, if in doubt, get legal advice.
• Link every gap you find to a cause and a concrete measure.

• Comparing metrics that are defined or calculated differently.
• Adopting industry averages with no source and setting them as targets.
• Presenting a tool's estimated visibility or traffic as a competitor's measured figures.
• Exchanging prices, costs or plans with competitors in order to compare.
• Publishing the results of a comparison with competitors without being able to back them up.
• Running the comparison and then changing nothing.