# Sampling to Predict Your Success

Last month I argued that detailed measuring of sales associate behavior and your sales process—not just financial outcomes—can mean the difference between a revenue increase of zero and a revenue increase of 10 to 20 percent. In addition, thorough measurement can provide an early predictor of future performance, allowing you to make midyear corrections.

## Appraising the Situation

I suggested measuring the following:

• The percent of customers you engage in conversation about their needs and dreams that results in your being able to recommend jewelry

• How effectively sales associates reengage customers who initially decline assistance

• How often your customers show an interest in an initial recommendation

• Where in your “sales funnel” the highest percentage of nonbuyers disengage from the sales process

• How often customers say “No, thank you”

• How customers perceive their sales experience in your store.

## Gem Bytes

Following are tips for getting this information.

Be realistic—think sample, not census.You need an indicator that shows you if things are improving, but don’t try to measure everything. Instead, use samples. To determine, for example, how often customers say “No, thank you,” randomly pick two dozen observation times over the course of one or two weeks and observe your sales associates’ interactions with 10 different customers. Note how often in each group of 10 observations you hear “No, thank you.” Take the mean of these numbers and divide by 10 to estimate the percent of customers who say that. Once you’ve changed your sales process to decrease the percentage, repeat the sampling effort to see if the percent has declined.

Randomness is key. Sampling in a random fashion is a requirement for making accurate projections to the population. Don’t confuse random with arbitrary. Arbitrarily picking times almost always introduces bias. Use a coin or die to help you draw a true random sample. For example, roll the die to determine which day of the week to make your next observations. Flip a coin to decide whether to observe in the morning or afternoon and flip it again to decide whether to observe in the first or second half of the time frame. Continue in this manner until you narrow down to a specific hour.

Be a multi-methodologist. Observation is just one method for collecting data. Some of my clients have created short check-lists on index cards for their sales associates to complete. Keep it simple. A card might include the date and time and the following questions:

• Customer wanted assistance immediately or later?

• Associate engaged customer in conversation to learn needs/dreams?

• Customer showed interest in early recommendations?

• Associate proved to the customer that recommendation met needs?

• Associate addressed questions/concerns?

• Associate suggested add-on item?

Give associates a coin and a die to develop their own random samples. To avoid bias, they must decide when to complete the card before speaking to the customer.

Ask your customers. The only way to know what your brand represents to others is to find out what customers think. Sample randomly and keep questions basic. For ex-ample, once a quarter you might ask 50 customers four or five key questions about their experience in your store. Ask orally or give them a small printed card to complete and drop in a box. Studies show most customers don’t mind being asked their opinion as long as it’s done in an easy, unobtrusive manner. Other studies show that, by letting them voice their complaint, you increase the probability a customer who has a complaint will return. If you can quickly resolve the problem, the chance of their returning is equal to or even greater than customers without complaints.

Years ago Lord Kelvin commented, “Unless you can say it with numbers, you can’t say anything meaningful about it.” Use the power of sampling to measure and understand your sales associates and sales process so you can improve it and more accurately predict your success.