Home
About Us
Our Services
Our Clients
Online Survey
Contact Us
ROI
 

 

 

We’ve done a lot of research in the past.  How is the ROI process different from customer research programs other companies have tried to sell us?

First, we build a strategic picture of the business from the “bottom-up.”  We integrate very detailed customer information with your financial data (current sales, profitability, and resource productivity) at the classification level.  This allows us to identify, monetize (express all the findings in currency and not just percentages) and prioritize your best opportunities – by customer segment, merchandise classification, price point, style, brand, store type, channel, region, etc.  For each classification – and in total – we quantify the amount of lost sales and profits and identify the reasons why regular shopping customers make their purchases elsewhere.  We differentiate between the “absolute” amount of lost sales and the “real” opportunity for our client at the merchandise classification level – where our client is at least in the “consideration set” among its customers. 

Second, our proprietary software system allows us to collect minute detail at the merchandise classification level.  We probe customers about individual transactions for the reasons they purchased elsewhere, the brands bought, prices paid, styles preferred (using digitized photographs selected by our client's buyers or style director), sizes worn, materials preferred and disliked, whether the purchase was a joint decision with a partner, whether a garment wore well, washed well, etc., etc.  We position many projects as “buyer led," as we interview key buyers and elicit their input “up-front” for the questions they would like to have answered.  Each merchandise classification can have a completely separate and detailed stream of follow-up questions, customized to the needs of each buyer.  By making the buying organization part of the process, the findings become much more actionable.  Similarly, we can add custom questions for particular stores, store types, or regions of the country. Retail is detail – and the ROI process provides information to buyers and operators that is immediately actionable. [Top Of Page]

 

For what kind of clients doesn’t the ROI Process work well?  Is there anything this approach is absolutely not good at/used for?

ROI's process works well for all kinds of retailers and direct merchants, though it has been most often used by large, established, multi-category store-based retailers.  The process is especially appropriate for fashion offers, but it has been very successfully applied to retailers of all kinds of hard goods.  Although it can be applied to food, supermarket retailers have less “room for maneuver” to edit categories and change the physical configuration of the stores than non-food retailers.  In general, though, the ROI Process is a powerful one that can be applied to retailers of all sizes and in all lines of trade.

Some suppliers can also benefit from the ROI Process. About two-thirds of all product purchase decisions are made by consumers while they are in the store, so gaining insight into customer shopping and purchasing behaviors can have huge impact. ROI offers a Category Scorecard™ product that is particularly useful to suppliers, and offers much more detailed information than they can gather from any other source. [Top Of Page]

 

Are there situations where the ROI Process doesn’t work?

The ROI Process is pragmatic and intuitive, but its success depends highly on the commitment of senior management.  When the process has been micro-managed by an internal market research department, it has often been doomed to failure from the start.  Often, the project will get bogged down into academic arguments like “should we use a five or six point Likert scale for measuring attitudinal data?”  These are the kinds of questions – from market research departments – that will kill the ROI process.  It must be driven by the highest levels within an organization for the process to be successful. [Top Of Page]

 

How can the approach be used to develop strategies? Reduce costs?

All the results of the survey are “monetized,” so it is easy for our clients to see where they have their greatest opportunities across their entire customer-facing business.  Senior management can immediately see what customer segments and categories offer the greatest real opportunities.  Operations executives can see what programs and services have the greatest influence on customers that offer the Company the greatest opportunities for profitable growth.  The ROI Process provides information to management that helps them allocate their scarce resources most effectively. [Top Of Page]

 

Why use a customer exit survey, rather than another type of data collection?

First, we focus on existing customers to drive the CIP process because it is the easiest, fastest, and least risky to gain more sales from customers who already shop the stores on a regular and repeat basis.  Our objective is to identify the barriers that prevent customers from buying more.  We want to “pick the low-hanging fruit.”

The customer survey taken on location at the store exit offers advantages not available in other types of surveys.  Exit intercept surveys can be used to acquire immediate reactions to a specific situation that has just occurred (and can measure customer service reactions and lost purchases while fresh in the customer’s mind) – in addition to collecting information that will define the customers’ characteristics, attitudes, and general purchasing behaviors.  Moreover, it is a relatively inexpensive method for acquiring in-depth interviews with a segment of the population whose number may not be large.  Locating these people by means of a market-wide or national survey might require considerable “screening.”  The on-location interview offers distinct cost and recall advantages.  All things being equal, the data collection should occur at, or near, the point of purchase -- and not in a market research office.

Mail surveys should generally be avoided because of the large non-response problem (plus, customers who care the most – either positively or negatively – are the most likely to respond).  With mail surveys, the respondent is far from the time and place of purchase.

In-home personal interviewing can be good from many perspectives, but it is expensive. Again, the store experience is not fresh in the minds of respondents. 

Telephone surveys, because of the coverage, control, and cost, also have their place, but don’t lend themselves as well to an ROI type questionnaire because it is more difficult to present photographic images to respondents via a phone survey and the customers’ shopping visit may not be fresh in their minds.

ROI can incorporate Web surveys as either a primary or secondary data collection vehicle. There are both advantages and disadvantages of using the Web, depending on the specific situation. Web surveys are fast and can be less expensive and less costly than exit intercept surveys. Once again, specific in-store experiences are less fresh in customers' minds.

In focus group testing, consumers are paid to come to a central office facility where they are exposed to concepts and interviewed. Focus groups also have their place, but they cannot provide quantitative data that can be statistically projected to the total customer base or market, and one or two strong personalities can easily sway the opinions of an entire group. [Top Of Page]

 

How can the process incorporate existing research?

In our experience, we have been able to consolidate many existing research programs into one, resulting in considerable savings to our clients through reducing their total research budgets.  For most of ROI’s long-term clients, in fact, we have been able to incorporate many individual research initiatives into one comprehensive work, providing a much richer, fully integrated marketing database.

ROI’s software permits many hundreds of questions to be included in the survey database, so other client research initiatives can usually be incorporated in their entirety – as a subset within the larger ROI questionnaire.  Obviously, we incorporate useful questions from prior customer and consumer surveys verbatim, so appropriate comparisons can be drawn. [Top Of Page]

 

How can the process be used to identify or validate brand strategies?

The ROI process is not Brand Research, per se, but it will provide valuable information for developing brand strategies.  “Brand Research” generically involves identifying relevant consumer segments (profiling them demographically, psychographically, and identifying their shopping and purchasing habits), and maps perceptions of the client company and its key competitors for the key segments.  Brand research usually focuses on “positioning.”  This type of research, as it is most often employed, is most effectively accomplished through a general market survey.

ROI’s process takes a very pragmatic view of opportunity space.  We believe – for any large, established retailer – that “repositioning” must begin by better serving current customers.  Our objective is to paint a comprehensive picture of current customers and identify opportunities to increase sales and profits from the customers who are already in the stores.  We learn the needs and wants of customer segments that offer the greatest opportunity by drilling down into the data, examining individual segments.  By tailoring the offer to meet the needs and wants of the target customer groups that offer the greatest potential, we believe new customers will automatically be attracted who also match the target customer profile.  We believe that “repositioning” strategies that do not start with regular and repeat shopping customers are risky, as the new “position” often tends to alienate customers who are the Company’s “bread and butter.”

The ROI process usually includes 80%-90% of the questions found in a Brand Research study, but our universe is current customers – not consumers in the total market.  We would employ market studies to provide context, but as a second priority. [Top Of Page]

 

Does the process accommodate multi-channel coverage? How?

In the simplist, most cost-effective scenario, a random sample of all customers exiting stores will include customers who buy some products from other channels (i.e. the Company’s catalog or Web site).  Streams of questions can be included in the questionnaire to address specific questions to these groups of customers.  The ROI process will quantify the cannibalization that is occurring between channels, and identify from which customers, and for what products. 

We most often administer the ROI process either via touchscreen kiosk (as an exit interview in a retail store environment) or through a telephone survey (for direct merchants).  We can also field an ROI survey on the Web, and the Web data collection can be either a primary or supplementary source for multi-channel shopping and purchasing information. [Top Of Page]

 

Can the process be used to test new products or collections?  Doesn’t this bastardize the core concept (e.g., existing customers, existing categories)?

The ROI process can accommodate questions about new products, but the core of the program is still about quantifying opportunities among existing customers in existing categories.  Photographs can be digitized or other physical stimuli can be provided to customers to gauge reaction and to ask specific questions about new products, which can later be related back to customer profiles and dollar opportunity. [Top Of Page]

 

Can we do a pilot test or use the ROI Process on just a few categories to save money?  What if we want to do only a few categories instead the whole assortment (because these few are our strategic priorities)?

Our process requires a certain “critical mass” to be successful.  Smaller scale studies with fewer respondents do not provide sufficient data for analysis. 

Our cost structure is relatively “inelastic,” and based primarily on fixed programming and other overhead costs.  Sample size, the number of stores involved, and the number and complexity of questions in the questionnaire have some bearing on our fees, but within a large relevant range, the incremental costs of adding categories, questions, or sample is low.  At the same time, the added benefit to applying the CIP across the range is high (detailed cross-purchasing information can be provided).

While pilot or smaller scale studies generally aren’t economically feasible, we can “un-bundle” our package price to try to come closer to a fixed budget constraint by providing “menu pricing” for some deliverables.

The ROI Process can be cost effectively applied to a Family of Business within a large company, and we have done that many times. [Top Of Page]

 

How long does the process take?

The “typical” full Opportunity Analysis project has an elapsed time of about 4 ½ months.  This timeframe is partially dependent on our client’s availability at key checkpoints. Less comprehensive Web surveys and Flash Reports can be delivered faster. [Top Of Page]

 

When is the best time of the year to do this?

Customer survey results may be biased by the timing of the interviewing (season, special events, time of day, etc.), which can affect the customer mix and quantity.  While periods of unusual events or promotional times should be avoided, there is never a “normal” time in retailing.  To await that period will mean that the project is never undertaken.

Having said that, the best times for the data collection phase of the project are late February to early June and early September through mid-November for a department store or other fashion retailer.  For home centers, the optimum data collection timing is typically May through September.  For some retailers, seasonality is not an issue.  It really depends on individual circumstances.

The ROI Process can also be set-up to provide continuous feedback to management, though this approach requires more custom programming and some permanent interfaces to our client's systems. [Top Of Page]

 

Can you do the process faster than indicated?  Does this negatively impact the output?

The entire ROI process has been completed in less than one month in exceptional circumstances (the client was in bankruptcy), but such radically expedited work may impact the detail of the output (there is not as much “think” time available for "drill-down" analysis), and it requires the prioritized attention and availability of client management. [Top Of Page]

 

How do you collect the data from customers about what they buy elsewhere?

Most often, our process for gathering customer information is based on temporarily installing touchscreen computer kiosks near the exits in a sample of stores.  We recommend that our clients use their own employees to recruit several thousand customers to participate in structured interviews, lasting just a few minutes each.  If permanent or temporary staff is unavailable, temporary services can supply the customer "recruiters." The whole in-store process typically covers an elapsed period of at least two weeks, and takes place during all days of the week and hours of the day the stores are open.

During the survey, customers are probed about the details of their purchases, as well as items they did not purchase and day and items purchased recently from other retailers. Information gathered from the survey is then merged with information in our clients' databases to form the complete picture of customers' purchases wherever they shop. [Top Of Page]

 

How do you cover a hundred merchandise categories and get answers to hundreds of questions if the survey lasts only a few minutes?

Our proprietary and patent-pending system allows many hundreds of questions to be included in a “typical” customer survey, while at the same time limiting the duration of the average interview to just a few minutes.  Because our software dynamically “adapts” the questionnaire script during the course of the interviews, we can make a hugely complex questionnaire appear short, simple and to-the-point to the customer – crucial for a successful exit interview.  Basically, three techniques are employed to shorten the questionnaire:

  • Conditional branching – Like a decision tree, software branches the respondent to the right question at the right time, asking only appropriate questions;

  • Individual questions have their own sampling scheme, and customers are chosen at random to be presented various questions, regardless of previous responses;

  • The software adapts over time the frequency of appearance of both questions and individual responses to questions.  This ensures statistical significance, while avoiding wasteful over-sampling.

In a recent UK customer survey, we fielded a questionnaire consisting of 639 questions.  It would simply be impossible to gather this level of detail through traditional market research means.  Our system provides the best of all worlds:  Customer data can be collected in the stores – while the shopping experience is fresh in the mind – and the detail collected is even more detailed and accurate than in-home surveys that could potentially take several hours to complete and require significant incentives to be paid to customers to participate. This same software technology can also apply to Web data collection. [Top Of Page]

 

Is it best to do all people in the store, or only ones who come in frequently?  How do you screen out (or should you?) tourists, seasonal shoppers, by-standers?

There are several schools of thought here, but we recommend recruiting a random sample of all “adult” customers (whether they purchased that day or not) as they exit the stores.  After the fact, we separate out frequent from infrequent customers, tourists, seasonal shoppers, and others.  This way, we have a statistically valid “snapshot” of the stores’ customer base. [Top Of Page]

 

What if the store doesn’t have much traffic?

We have successfully used this process for collecting customer data in stores as small as a few hundred square feet to as large as one million square feet.  We have completed projects in fashion stores, in hardline stores, in stores with very little traffic (a handful of customers per day), and in stores with tremendous traffic (over 100,000 customers per day).  The worst case – in stores with very little traffic – is that the survey must take place over a longer time period to achieve a sufficient sample of customers for further analysis.

The touchscreen terminals are not intended for use by your customers, unaided.  We normally train our client’s salespeople to function as recruiters – and let the customer complete the interview only if he or she is clearly comfortable using the technology.  Even in that case, the salesperson must make herself available to answer any questions. Very low volume stores must have enough personnel available to man the terminals during busy times, so it may be necessary to dedicate temporary personnel to this function. [Top Of Page]

 

For a large national specialty chain (500 or more stores), how many stores would typically be surveyed? What factors do you consider when determining how many stores should be surveyed?

The number of stores to be surveyed depends on a variety of factors, including, but not limited to, the commonality of the store sizes and configurations, their geographic distribution, and competition.  We try to include a cross-section of all types of stores in the chain, so we can compare and contrast the profiles of successful stores with relatively unsuccessful ones, large stores with small ones, stores in metro markets and rural markets, stores in regional malls, strip malls, and free-standing stores, stores in markets with heavy competition versus those with light competition, new stores versus old stores, etc.  We draw up a matrix, with management, to ensure we have a representative sample of stores covering the most important combinations and permutations within the chain.

As a general rule of thumb, we start with a store sample of about 10% of the total number in the chain. For chains with a large number of similar type stores, we usually find we can cover the different combinations of interest with a sample of about 20-30 stores.  For chains of larger department stores, where there are often distinct differences between the stores or customer bases (but fewer stores), we often include all the stores in the chain.  Obviously, the number in the store sample depends on the individual circumstances. [Top Of Page]

 

How do you minimize store disruption?

Terminals are located near the store exits, but out of the main flow of traffic.  A small table and two-three chairs are required.  There is virtually no disruption.  Nevertheless, the store manager must consider the survey a top priority during the two-plus weeks it is occurring. [Top Of Page]

 

Don’t customers find the process intimidating?  Doesn’t this have the potential to tarnish my store image (intercepts at doors)?

The intercepts are discrete, associates are trained to follow tested methods upon approaching customers, and the interviewing table is not in the flow of traffic.  We’ve found customers legitimately enjoy the experience and the opportunity to express their opinions.  We have never had any complaints from management following the completion of the survey, and in fact, we have had many positive remarks about being able to listen to customer feedback and suggestions.  This has been the case in all kinds of stores, from discount to high-end specialty department stores. [Top Of Page]

 

Should we provide a customer incentive?

Incentives are always a nice touch, but are not strictly necessary for the successful completion of the survey.  If management decides to provide an incentive, we recommend our clients provide a small gift, a discount coupon redeemable on the customer’s next visit, or have some sort of drawing for a valuable prize, like a television. [Top Of Page]

 

How do you quantify “style” preferences?

We quantify the style preferences of your customers at the classification level by digitizing photographic images of different kinds of merchandise selected by your own designers, buyers and merchants (not just the styles you currently carry), load them onto our survey computers, then show them to customers during the interviews.  Everyone – designers, buyers, merchants, management, researchers, and customers – finally talk the same language! The client should probably consider assigning an appropriate buying oriented manager the task of coordinating this effort (it will take a few days).  This may be the project manager, depending on the skills of the manager. [Top Of Page]

 

We are too fashion-driven for this approach (what sells is what’s “hot”).  How can this help me?

Parts of the outputs of the ROI process are strategic and parts are very tactical, but the process can never replace the judgment of a good buyer.  It can provide information to allow the buyer to work smarter, with more in-depth information about customers and their specific buying patterns, but it cannot predict the “next big thing.”  It can quantify opportunities in various categories, provide specific information about price points and customer sizes that are under-penetrated, and can offer guidance on customer style preferences.  Buyers must incorporate this information into their buying plans, along with their judgment about styles and features that will be “hot” next season. [Top Of Page]

 

Who needs to be involved from our organization?

Even with a project coordinator, for the results of our work to be most actionable, we want to work closely, up-front, with the management, marketing, and buying teams.  The up-front involvement of the buying/design team ensures a high level of organizational “buy-in.”  We have found people are much more likely to implement the findings if they are involved in the design of the program and the content of the questions. [Top Of Page]

 

How much time is required up-front of individuals?  At what point does senior management get involved?

To kick the project off, there should be a series of one-on-one meetings – of around an hour each – with the CEO, the head of marketing, and the head of buying.  Although one hour should allow ample time, about 1½ hours should be allotted between appointments, so there is some flex in the schedule if a meeting runs long.  These meetings are informal, information gathering discussions, loosely based on the following topics:

  • Review of ROI process

  • Project overview, objectives, and expectations

  • Discussion of client strengths, weaknesses, opportunities, and threats

 

Normally, brief (less than one-half hour) meetings are also scheduled with the head of finance to discuss financial reporting requirements, as well as with someone from the IT department to discuss the nature of data extracts.

A group presentation is often made to the buying staff – or key members of the buying staff.  For a fashion retailer, someone from the buying group should be assigned the responsibility of pulling together images for buyer review and approval (if style images are used).  This may or may not be the project manager, depending on the skill set of the client-side project manager.

Later, someone from senior management should review the questionnaire prior to fielding, and the entire senior management committee should be present for the final presentation of results (usually a half-day session). [Top Of Page]

 

Will it disrupt their ability to do their regular jobs? 

Involvement in the ROI process consumes only a few hours of time over a 4 ½ month period for all but the project manager and the store staff that performs the interviews. While the interviews are going on in stores, the terminals must always be manned. The workload directly related to this program typically consumes no more than 10%-15% of the Program Manager’s time over the duration of this project (with most of the time required within the first month and immediately prior to survey fieldwork). [Top Of Page]

 

Do you need assigned individuals to the project? Who? What role?

ROI requests the client nominate a marketing/buying oriented associate to act as Program Manager and as a full working partner of the ROI team.  We have also worked with the Vice President of Human Resources in this capacity, with excellent results. [Top Of Page]

 

Who all on the client side needs to be involved, what is their role, at what points are they involved, for how much time?

Besides the project manager, sufficient store selling staff (or temporary personnel) are required for the duration of the surveys to administer the interviews.  In order to minimize the impact of external, environmental influences, we recommend that interviewing take place over an elapsed time of at least two weeks.  Senior management needs to be involved at the very beginning of the process and at the final presentation. [Top Of Page]

 

What are the key milestones?  Key QC/checkpoints?

See the Timeline Flowchart in the ROI Process slide show. [Top Of Page]

 

What are the deliverables of the process?

Our deliverables are tailored to the needs of those who will use the data.  Buyers get information specific to their categories.  Executives receive a summary of findings and recommendations and a full reference report. 

For management and marketing, all survey questions are monetized in terms of dollar opportunity, then organized into eleven report sections, useful for strategic decision making:

  • Executive overview, key findings and recommendations

  • Core customer description, shopping, and purchasing behaviors

  • Core merchandise, including categories that should be emphasized and de-emphasized

  • Core competition – in total, and at the classification level

  • Key appeals – what brings customers into the stores

  • Merchandise character, styles, brands, and perceptions of quality

  • Pricing

  • Promotion and advertising effectiveness

  • Customer service and services provided

  • Physical facilities and in-store ambiance

  • Appendix

 

For example, all pricing related questions are organized together, with a brief summary at the beginning of the section.

We run a variety of statistical analyses on the data and present information in the reference report, highlighting exceptions.

Sometimes, we provide “Category Scorecards™ ,” covering all classifications a buyer is responsible for.  This category summary report details opportunity (in dollars) from different customer segments, it quantifies lost sales (again, in dollars) to each significant competitor, identifies the reasons customers buy at the client and buy elsewhere, the prices paid elsewhere, etc.

Though we provide the actionable detail at the classification level, we are very much conscious of identifying the “Quick Wins” for management – those areas of the business that offer the greatest return from the smallest investment in shortest possible time.  [Top Of Page]

 

What sorts of sales increases and other benefits can we expect from undertaking CIP?

The way we measure share of wallet, a well-established department store client will typically earn between a 25% and 50% share of its regular and repeat shopping customers’ spending for the kinds of items it sells.  If a retailer has gross sales of $1 billion, this fact suggests their customers spend, in the aggregate, a total of about $2 to $4 billion at any store in the markets it serves, leaving a total unrealized opportunity of $1 to $3 billion.  Clearly, this is a huge potential.  Just as clearly, though, no retailer can earn 100% of its customers’ purchases, and we try to highlight the “real” opportunities, where the client is in customers’ consideration sets and then prioritize those opportunities.

Virtually all major retailers manage their businesses based primarily on sales performance in previous periods, so no matter how well managed they are, our customer-focused process will always uncover inconsistencies in their offers.  Addressing these inconsistencies (in price, style, size, or in-stock position, for example), and making your offer more consistent with the needs and expectations of your target customers, will gain you a number of “Quick Wins.”  To put this in perspective, for a well managed $1 billion retailer, based on our experience with other retailers (and assuming ROI’s information is actioned), we have seen net sales gains of approximately 2.5% to 20%. [Top Of Page]

 

How do you calculate project fees? 

Our system does not fit a standard market research financial model, and we generally do not base professional fees on a “per interview” basis.  Projects are quoted on the basis of fixed professional fees, plus expenses.  If our client elects store exit interviews for primary data collection (rather than Web surveys, for example), we can provide hardware required for data collection if the client does not have touchscreen terminals available for this purpose. [Top Of Page]

 

Can I change the approach/output to reduce the cost?

The process is rather inflexible, but some outputs can be eliminated to reduce the total cost. [Top Of Page]

 

Are you prepared to work on a “success fee” basis?

Yes.  We are confident we will uncover a number of “Quick Wins” for our clients, and we are prepared to offer our services on a “success fee” basis.  As a sign of good faith, we ask the client to pay for project expenses only, and we will provide all services at no fixed fees, but instead for a percentage of your incremental sales over and above your sales plan over a three-year period.   We base this arrangement on a three year timeframe because there will be a time-lag before many changes can be implemented.  As a result of gaining a deeper understanding of the customer and her shopping and purchasing behaviors, we believe you will reap long-term benefits beyond initial successes. [Top Of Page]

 

What is most important to ensure the success of this approach?

By far the most important element is the commitment of senior management.  Following the completion of the project, programs must be installed to ensure buying plans and other programs take into account the findings of the project, and these must be monitored by management. [Top Of Page]

 

Is the ROI Process a continuous process, or does it present a picture at a point in time?
 
Typically, the ROI Process takes a snapshot of the business at a point in time, which can be repeated in multiple waves.  The ROI methodology is ideally suited to continuous tracking, but that would require permanent I.T. links to feed client financial, transaction, and customer data to ROI. [Top Of Page]

 

How long before the data goes “stale?”

Some elements of the data change rapidly, depending on market conditions and competitive actions.  Most of the information, though, has a relatively long shelf-life.  For a large, established retailer, customer characteristics and attitudes change slowly, and most elements of the research have a useful life of up to two-three years. [Top Of Page]

 

Should the process be repeated?  How often?

With long-term clients, we have applied our “traditional” Opportunity Analysis approach, then determined which parts should be repeated seasonally, yearly, or every two-three years to track performance and changes in the market.  Most clients find it useful to follow-up every one to three years.  If the company is undergoing radical change or repositioning, then more frequent follow-up is appropriate. [Top Of Page]

 
WEBMASTER LEGAL DISCLAIMER
©2007 RETAIL OPTIMIZATION INTERNATIONAL, INC.