The Challenge of International Market Pricing


Chuck Csizmar – CMC Compensation Group

“What is the competitive market price for a particular position?”

It’s a simple question.  If you work in Compensation, this is what you do.  And if you’re in the US, the survey sources you can call upon are numerous and well-stocked with participating companies and benchmark matches – the blessings of a large country.  In fact, it is a common practice to segment the data (report separately) on the basis of industry, revenue size, or geographic region.  In some instances you can further refine your analysis by operating budget, staff size or even years of experience.

For those accustomed to such robust analysis it can be a real wake-up call when asked to conduct a similar analysis for operations in another country.  Suddenly your content-rich environment has disappeared, and in its place you find that the availability of good information can no longer be taken for granted.  Now what do you do?

Your large country database is gone.  Instead, you face a limited selection of survey sources and each offers only a fraction of your normal participant count – a far cry from business as usual.

Such is the key challenge when pricing international jobs – the limited number of companies included in surveys, even by the major vendors.  For example, Mercer Netherlands has 81 participating companies.  So it is not unusual for a market pricing analysis to include only 4 – 5 “matches” – but is that representative of common practice?

If you’re the one on the asking end of the original question, let me share the challenges your analyst is likely to encounter.

Impact of Reduced Participation

  • Limited industry segmentation:  Reported data will likely cover multiple industries, with limited or no segmentation.  If you’re in either a high or low paying industry, surveys will provide inflated or discounted  information.
  • Hard to segment by revenue size:  To the extent that larger companies pay more than smaller you lose that distinction as well.  This can be especially problematic if you’re a small company.
  • Global responsibilities vs. strictly national:  The distinction is often blurred between national, regional and global responsibilities.
  • Combination jobs not well represented:  You will find yourself matching against jobs “close to” your own, just to gain a “feel” for pay levels.  If your job content varies from benchmark descriptions, reported data might not capture such idiosyncrasies.
  • Poor matches and / or no data when less than 5 respondents:  Surveys tend to provide an “n/a” when they do not have enough participants.  When you start with limited companies it’s not unusual to find unreported jobs.
  • Forget Regional variations:  While it is often the case that certain geographic regions have higher pay levels, the reported data is usually national.  You may assume that participants are in the higher paid region, at your risk.

What to do?

Frustrating, isn’t it?  You can’t very well throw your hands into the air, complain about poor survey quality and move on to something else.  The limitations are there and you have to play with the cards you’ve been dealt. Management is waiting, wondering what is taking you so long.

Working with limited resources is a test.  Your challenge is to balance an understanding of the subject position, the industry and the vagaries of limited data points in order to determine which figure best represents your position’s competitive value.

To succeed you must utilize subjectivity and your professional judgment to consider the available data and gauge which figures best reflect the job under review.  The correct answer will no longer jump off the page at you.  Compensation has become an art, not a science.

  • To improve your matching, consider either the 25th or the 75th percentiles instead of the median or 50th percentile to reflect your position: this can be effective with poor matches, or concerns that the reported job is either larger or smaller than your own.
  • You may have to add or subtract from a benchmark job to gain a more appropriate figure for your position.  For example, if your job is a VP but the survey matches stop at the Director level (or converse), you may have to adjust up or down to create a better “guesstimate.”  Note: in such a case don’t forget that the incentive percentages will likely differ as well.
  • There is no formula in making adjustments, but changes in organizational level are usually around 15% – 20%.  Within-level description changes are usually around 5% – 15%.
  • If dealing with only a few positions you might have greater success by individually pricing jobs through a vendor’s database of multiple surveys, government sources and local surveys.  Vendors like ORC, Birches Group and a few others offer this select service.
  • Be careful of the arithmetic exercise (averaging averages, inappropriate matches, assuming numbers, etc.) that delivers a figure you cannot validate later.  Caution: a number is remembered, while often the qualifiers that follow are forgotten.  Make sure that you document such concerns before providing specific data.

All this subjectivity means that your judgment might suffer from more skepticism, even criticism, as you cannot simply point to a survey page and say, “there it is.”

Does all this subjectivity ruin the value of your analysis?  Not at all, as long as you inform management about how limited survey resources have impacted your analysis.  They expect an answer to their question (market value?) and you need do the best that you can with the resources you have available.

More About Chuck:

3 responses to “The Challenge of International Market Pricing

  1. Chuck,

    This is an excellent topic, and well presented.

    In our experience in developing countries, many of the points you make apply even more — there simply aren’t enough large, well-structured employers in many markets to establish a sector survey, and sometimes it’s tough to get a multi-sector one. The dynamics of a small to medium-sized developing country are different from western Europe or America; job definitions are more fluid, there are many more incumbent-based job profiles and the size of the workforce can vary dramatically from company to company.

    Birches Group uses multi-sector data with cross-occupational job matching, grouping positions of equal contribution levels together, to ensure sufficient data is available to report back in the survey. This method works extremely well, and reflects the realities of these country markets very well.

    My advice echoes yours — be flexible, and adapt your approach. Remember, compensation is part art and part science :-).


  2. Chris Foster

    Chuck, this article is timely as we are once again participating in market surveys and performing TTC analysis in Europe. I have found that it is either feast or famine. Some countries have solid data where I can benchmark 80% of our jobs, others country surveys are considered “light” in that only 5 jobs are represented. I struggle with obtaining solid Sales Job TTC data and benchmark jobs – any suggestions here would be appreciated. Another functional area that needs improvement is the manufacturing production jobs. All production type jobs seem to be clustered under the “production” or “forklift” job which is not representative of complex machine operations as an example. I agree with you there isn’t a clear separation between Senior Manager and Director or Director and VP.

    What is more surprising to me is global companies have moved to or are moving to a global job leveling approach where you have a standardized job code/job title scheme across the corporation and where many of the major survey vendors have not adopted this approach. Continually, we need to match our standardized job titles to each Region of the world rather than just one time per vendor. Mercer does offer this Universal job scheme, but as Compensation departments get thinner and thinner, utilization of this model by vendors needs to be adopted not only for efficiency, but for consistency and reliability. This lack of standardization combined with lack of the basic levels of a job, (e.g., Entry, Intermediate, Senior, Senior Specialist, Master) forces increased slotting. I would suggest our vendors adopt the model companies have moved to – Universal job code scheme and consistent levels for each position. In the long run, this will save everyone time, ensure accurate survey participation and benchmarking, and limit the need to add a premium, increased weighting or use the 25% or 75% of particular position.

    • Chris –

      You have hit on one of the biggest issues with salary surveys available in the market today — the heavy reliance by the consultants on their clients to properly match jobs in the survey submission. Think about the process — a specialized methodology, developed by highly trained consultants, is adopted for the survey matching. While consultants conduct job matching sessions (usually in groups), in the end, it’s the clients, not the consultant, who are responsible for submitting the matches. Now, you may have someone on your team who used to work for consultant X, and knows their system really well. But what about the other participants in the survey? Not everyone is expert! And if you participate in multiple surveys, well, you get the picture.

      The solution is to turn the survey data collection process upside down, which is what we have done at Birches Group. Our analysts interview each employer in the survey, and based on the interview, the analyst matches the jobs. It’s consistent, we control the quality and it ensures our surveys are reliable.

      The impact of global grading is spot on as well. We have developing mappings from the global grading structures of over 25 leading international employers to the Birches Group levels, which are used to level jobs in our surveys. We still confirm them during the interview, but it serves as a starting point and ensures a level of consistency for the employer across all of our surveys.

      Now the real question, is why haven’t the “majors” figured this out?