In the many seminars I have done, including one for small businesses just a couple of week ago, it is clear that many businesses will benefit from increasing their use of tests when making decisions on whether to change their marketing mix. Because they do not test enough, they delay taking action on possibly good ideas and so pay a price in slower growth or smaller profit.

A way to get a grip on whether a possible move (e.g. pricing, additional item, different way to display important items, different packaging, etc.) can make sense it to test it in a limited part of your operation. This could be done by giving say 10% of your salesmen a different offer, or simply a different way to pitch an offer. Or if you have 10 food carts selling the same range, you can try something different in 3 of them. 
Small tests like these will not give you all the answers but you will learn enough to reduce your decision risks.
They can also highlight operational issues that were not anticipated.

There are a few things to keep in mind when testing. First, you should have enough salesmen or outlets where you will test. This is to make sure that your test will be done among enough salesmen or areas so that if you average their performance, that performance is reasonably close to the average performance of your whole organization. Close enough is roughly + or – 10% as a general rule of thumb. You may have to change some of your initial selection to achieve this.

When testing, it is usually helpful to have some salesmen or some outlets or some days which can serve as a reference point for comparison of the performance you get with the salesmen, the outlets or the days when you do your test. The reference salesmen or outlets or days are what are called your, “control”.

Before you start your test, make an effort to compare the sales performance of your test areas or people versus your control areas or people. For example, look at the average I mentioned earlier and try to have test and control averages which are within 10% of each other.  Again you may have to change your initial selection to remove those which are too unusual.

Then you also compare the monthly trend, at least over the past 6 months for your test versus control. If you have the data, compare monthly sales (or quarterly sales if your numbers bounce around a lot) for at least this year versus last year or a minimum of 18 months. One convenient way to assess this, is to do a bar graph which shows the performance of your total test and your total control salesmen. And check how they look. Red flags are times when one group’s performance spikes while the other group stays flat or even reverses direction. That suggests that different factors influence performance in your test versus your control. If so, then look for another set of control outlets, or test outlets, depending on which is easier to find.

Once you get the same basic pattern of ups and downs in both areas, you can then check the general growth trend. Look at the growth percentage for your test outlets for say, year-to-date, this year versus the same period last year and do the same for your control. You want to have percentages that are in the same direction (both up or both down) and in percentage terms within 5% of each other.

Your test period (when you try things) could be as short as a week if you believe from experience that you already will see meaningful changes in that span. But I would take into account the importance of the decision to you (e.g. the cost to roll it out to all stores, the damage if something goes wrong) and suggest a somewhat longer period before deciding.  Now go ahead and try testing. Good luck!



Benedicto “Poch” Cid is the Chief Brand Adviser of Mansmith and Fielders, Inc. (www.mansmith.net), the leading marketing and sales training company in the Philippines. He will be conducting the seminars entitled the 11th Pricing Strategies and Tactics on November 27-28, 2014 and the 12th Brand Management: Understanding DNA and Engagement on December 9-10, 2014.  For inquiries, please email info@mansmith.net, call (+63-2) 584-5858 /412-0034 or text (63) 918-81-168-88. You can also send your marketing, sales and strategy questions to mentors@mansmith.net