Groupon – Show Me the Money.

Groupon has shown the world a new way to do coupon promotions, but is it any better than traditional couponing?  Is Bizy Deal a good deal for suppliers?  In this article, provide an academically proven and industry best-practice approach for evaluating coupon promotions, be it Groupon or any other type of couponing.

Development of Updated Examination

The Wiglaf Journal has looked at Groupon in the past.  In Groupon: Goldmine, Tar Pit, or Niche Solution, we examined the economics of a Groupon from the firm’s perspective and argued that Groupon addresses a niche challenge in the promotional industry.  Shortly thereafter, a New York Times Boss Blog posting leveraged this approach with a very similar argument in Doing the Math on a Groupon Deal.  Then, we explored the potential for this firm to thrive as a stand-alone entity in The Groupon Phenomenon — Is It Sustainable?  Each of these postings received a strong response. Clearly, we touched upon an important issue.

Our subsequent discussions with executives indicated that many struggle to evaluate the effectiveness of a coupon, much less a Groupon promotion.  Executives understand that the purpose of couponing is to drive customer acquisition, but how should they measure the effectiveness of a coupon / Groupon?  Total costs of couponing?  Total coupon related sales?  Redemption rates?  Something else altogether?  Or, perhaps a hybrid metric?

In preparing prior posts, we proposed promotion metrics developed from the work of Scott Neslin, the Albert Wesley Frey Professor of Marketing at the Amos Tuck School of Business Administration, Dartmouth College.  After posting an early version of a model of coupon effectiveness, I had the pleasure of corresponding with Dr. Neslin directly regarding the calculations.  Our discussions led to a clearer understanding of the appropriate modeling – specifically as it relates to Groupon.

Based upon the response to prior posts on this subject, we suspect executives would like to know the results of our discussions.

Warning for readers:  this is a technical analysis of the effectiveness of Groupon and other forms of couponing.  If you want to know how to evaluate promotions, this article will provide a template.  It is written to provide a resource text to executives seeking to understand best practices in promotional management.

Analyzing Promotions

Analyzing the effectiveness of coupon promotions requires tracking redemption rates, the total campaign cost of the promotion, the campaign profits before promotional costs, and the percentage of coupon redeemers that are incremental customers, customers that wouldn’t have purchased without the coupon.

Once these metrics have been measured or calculated, a model can be used to evaluate the campaign effectiveness.

To demonstrate, we have prepared the attached spreadsheet on Wiglaf Pricing Explores Groupon v Newspaper for worked examples of the analysis and model as applied to Spring Restaurant, a reference client for Groupon.  The model for Spring Restaurant’s Groupon experience is provided in column C.  That for a hypothetical but relatively comparable coupon distributed via the newspaper is provided in columns E.

Download the spreadsheet Wiglaf Pricing Explores Groupon v Newspaper

Redemption Rates

Despite their value, most coupons are not redeemed.  Even Groupons, which are coupons that customers pay money to receive, are not all redeemed.  Some are simply forfeited.

For Groupon, redemption rates vary between firms but reports of redemption rates as low as 70% have been identified.  That would imply that a whopping three out of ten Groupons go unused.  These reports might be at the extreme end of a poor response rate, hence we will assume in our Spring Restaurant example a slightly higher 80% redemption rate.  (Managers are encouraged use their own expectations in constructing their model.)

For coupons distributed via traditional means, the redemption rate is significantly lower.  In 2010, consumer packaged goods firms distributed 332 billion coupons via free-standing inserts, in-store media, digitally, and through other means.  Of these coupons, only 3.3 billion were redeemed.  Thus, on average, a manufacturer of a commonly purchased good should only expect 1% of the coupons distributed via traditional media to be redeemed.  A seller of a specialized good, such as a restaurant experience, should expect a significantly lower redemption rate.

Spring Restaurant reported sales of 2,873 Groupons when they ran their promotion.  If 80% of these Groupons were redeemed, then 2,298 tabs would have been paid, or partially paid, with a Groupon.  (See cells C15, C30, and C31.)

For a comparable alternative, Spring Restaurant could have run a targeted coupon in the Chicago Sun Times, Friday – Weekend Edition, North Side Distribution Area.  This would have reached approximately 75,000 readers.  (The Chicago Sun Times reports a distribution run of 312,655 in their media kit.  A targeted advertisement run on the North Side is reported to reach 24% of their readers, or roughly 75,000 readers.  See cells E19-E21.)

Of those coupons distributed via the Chicago Sun Times, Spring Restaurant should expect redemption rates significantly below the industry average for consumer packaged goods because restaurant experiences at unique upper-end restaurants are targeting a smaller customer base and purchased significantly less frequently than consumer packaged goods.  If the average redemption rate for consumer packaged goods is 1%, it is reasonable to assume the redemption rate for Spring Restaurant coupons distributed via the newspaper may be ten times lower, or only one-tenth of a percent.  That would imply a single, targeted coupon run in the Chicago Sun Times might be expected to bring in 75 new customers.  (See cell E21, E30, and E31.)

Campaign Costs

Total campaign costs of couponing include both the redemption costs and the distribution costs.

Redemption costs are the costs associated with redeeming a coupon.  This includes the face value of the coupon and the costs of processing coupon redemptions.

• For the Spring Restaurant example in the spreadsheet, the Groupon deal offered customers a $75 dining experience for a $35 Groupon – netting the customers $40 in savings.  (See cells C9-C11).  When the customers visit Spring Restaurant, their Groupon is worth $75 of food and beverages.  Thus, the face value of the Groupon for Spring Restaurant would be $75.  If 2,298 Groupons were redeemed in this campaign, the total redemption costs would be $172,380 (=$75 X 2,298). (See cell C32 from cells C31 and C9.)
• As a comparable alternative, Spring Restaurant could have run a $40 (=$75 -$35) coupon in the Chicago Sun Times (Cell E18) which provides customers with the same level of discount while not forcing them to pay upfront for a coupon which they may not use.  If 75 customers redeem this coupon, the redemption costs would only be $3000 (75 X $40).  (See cell E32 from cells E31 and E18)
• Coupon processing costs are zero in our Spring Restaurant example as it is assumed that Spring Restaurant managers process their own coupons.  However, manufacturers coupons redeemed at a grocery store or other retail outlet are often processed by a third party redemption service for a small fee; hence these costs would be included in a model of promotions for many other businesses.

Distribution costs are the costs incurred to distribute the coupons to prospective customers.

• With Groupon, the distribution costs are zero and actually there are distribution revenues.  This dramatically alters the economics of a Groupon.  Groupons are sold to customers for them to redeem.  As a baseline expectation, Groupon shares 50% of the Groupon sales revenue with the advertiser, and pockets the remaining 50%.  Hence, a Groupon with a face value of $75 sold to a customer for $35 nets the advertiser an instant $17.50 (=50% X $35).  (See cell C14 and derivation thereof in cells C10-C13.)  If 2,873 Groupons were distributed in this campaign, the revenue from Groupon sales would be $50,277.50 (=2, 873 X $17.50). (See cell C34 from cells C15 and C14.)  This Groupon distribution revenue will offset a portion of the redemption costs in calculating the total campaign costs.
• If Spring Restaurant had run a coupon in the Chicago Sun Times, Weekend Edition, North Side Distribution Area, the distribution cost would have been approximately $948 to insert the advertisement.  (See cell E44 and derivation thereof in cells E22-E26.)  The distribution costs associated with other traditional couponing techniques, such as direct mail, leafleting, in-store distribution, etc., should be treated in a similar manner as that shown for the newspaper, though the calculation of the distribution costs would differ.

Total Campaign Costs are the sum of the redemption costs and distribution costs.

• For the Groupon campaign described, the total campaign costs would be $122,102.50, the coupon redemption costs less the coupon distribution revenue (=172,390.00 – $50,277).  (See cell C35.)
• For the newspaper campaign described, the total campaign costs would be $3,948, the coupon redemption costs plus the coupon distribution cost (=$3000 + $948).  (See cell E35.)

Campaign Profits

Campaign profits are the profits generated from the sale of goods prior to including the campaign costs.

For Spring Restaurant, the average tab per person is $59 according to Zagat.  Assuming that people mostly eat at fine dining restaurants in couples or larger groups, the average bill per table would be $118 or higher.  (Cell C4 and C5.)

Coupon users however are not the average diner.  As a market segment, it is strongly suspected that couponers have a lower average dinner tab, tip less, and are generally frugal.  These factors may lower the average tab associated with coupon redeemers from that of the average restaurant customer.

Without actually seeing the receipts associated with Groupon or more generally coupon users in comparison to average customers, I can only speculate on the differences between these market segments.  As a default rule, I will assume the null hypothesis until sufficient evidence proves otherwise – all customers are the same.

Thus, for modeling purposes, we will assume an average tab associated with coupon redemption to be that of an average couple dining at the restaurant, or $118.  (Managers are encouraged use their own expectations in constructing their model.)

For our model, gross margins at restaurants have been estimated to be near 60%.  Much of the cost of running a restaurant is relatively fixed, such as rent, labor, and overhead.  Little of the cost of running a restaurant is in pure food costs.  In evaluating promotions or other forms of discounting, a strict application of pure marginal costing has been advocated by many economists and pricers.  If only pure marginal costs were used, it is expected that the gross margins are higher.  We have chosen a gross margin of 60% in our model based on conversations with restaurateurs and accept that different restaurants have different gross margins.  (Cell C6.)  (Again, managers are encouraged use their own expectations in constructing their model.)

For Spring Restaurant running a Groupon campaign, we therefore calculate revenues of $271,211.20 to be generated from the campaign (=$118 per table X 2,298 Groupons redeemed). Similarly, we calculate the cost of goods sold to be $108,484.48 (=$271,211.20 in revenue X 1 – 60% gross margin).  Thus the total Groupon campaign profits are $162,726.72 (=$271,211.20 in revenue – $108,484.48 in cost of goods sold).  (See cells C37-C39 from cells C5, C6, and C31.)

In comparison, if Spring Restaurant had run the newspaper campaign described, the campaign would have yield $8,850 in revenue (=$118 per table X 75 coupons redeemed) against $3,540 in cost of goods sold (=$8,850 in revenue X 1 – 60% gross margin) yielding a total coupon campaign profit of $5,310 ($8,850 in revenue – $3,540 in cost of goods sold).  (See cells E37-E39 from cells C5, C6, and E31.)

Modeling Effectiveness

From the above analysis, it is clear that redemption rates influence the effectiveness of a promotional campaign.  However, the above discussion does not clarify how they affect campaign efficiency.  In fact, under certain circumstances a campaign with a lower redemption rate can be more efficient than one with a high redemption rate, and vice versa.  Let use demonstrate.

There are three core metrics for evaluating promotional campaign efficiency:  (1) Cost per buyer, (2) ROI, and (3) BISR.

The cost per buyer defines the customer acquisition costs.  Analytically, this is simply the total campaign costs divided by the number of coupons redeemed.

• If the product is a consumable good or frequently purchased consumer good, firms may be willing to “subsidize” the first purchase if the firm expects that first-time customers will become repeat purchasers.  Repeat purchases often drive the bulk of a firm’s profits and the majority of the full customer lifetime value.  Cost per buyer metrics can be used in modeling the total customer lifetime value. For firms selling to markets where a bulk of the revenue is derived from repeat purchases, the cost per buyer is an important metric.
• If the product is a durable good or infrequently purchased consumer good, the cost per buyer holds little information for evaluating campaign effectiveness though it can be useful for budgeting purposes.
• For Spring Restaurant’s Groupon campaign, we calculate the cost per buyer at $53.13.  Similarly, for the hypothetical newspaper coupon campaign, we calculate the cost per buyer at $52.64.  (See cells C36 and E36.)  Thus, either of these couponing approaches cost Spring Restaurant approximately $50 to acquire a new customer.

The ROI (return on investment) is simply the gain less the pain divided by the pain, or more analytically stated, is the ratio of the campaign profits less campaign costs to campaign costs.

• If the campaign has a positive ROI, the firm can conclude the campaign may be profitable.  If the campaign has a negative ROI, the firm can conclude the campaign is definitely unprofitable. As a first line of campaign effectiveness analysis, campaign ROI should be used to segregate profit destroying campaigns from potentially profit enhancing campaigns.
• We say “potentially profitable” because this approach to calculating ROI is really glossing over an issue.  It is effectively assuming that the coupon redeemers wouldn’t have visited the establishment without the coupon.  Surely, at least a portion of the coupon redeemers would have purchase anyway in the absence of the coupon.  As such, this approach overstates campaign ROI.  (In other words, yes, the company made money, but it could have made more money if the coupon redeemers had purchased without a coupon.)  We accept this challenge and suggest that it be handled in the next campaign metric.  Most firms calculate campaign ROI in the approach mentioned.  We agree this approach can be misleading, and as such suggest firms invest in understanding an applying the third campaign metric, the BISR.
• For Spring Restaurant’s Groupon campaign, we calculate the ROI at 33%.  Similarly, for the hypothetical newspaper coupon campaign, we calculate ROI at 34%.  (See cells C40 and E40.)  Thus, for either of these couponing approaches, every $1 invested in couponing appears to return over $1.30 in profit, and thus they appear to be a good deal – but this is slightly misleading as we will see.

The BISR (Breakeven Incremental Sales per Redemption) is the true key metric to examine.  It defines a hurdle which the campaign must exceed before it is declared effective.  Specifically, it defines the number of customers which must be incremental, that is customers that would not have purchased in the absence of the campaign, for the campaign to be declared profitable.

• The BISR highlights the need for a firm to distinguish deal customers between those that are incremental and those that are not, even though both are deal customers.  Some deal customers are new customers or repeat customers that are increasing their spend.  These deal customers would be identified as incremental customers.  Other deal customers are customers that would have purchased anyway but are simply taking advantage of a good coupon.  These are non-incremental customers.  Firms need to track or estimate the ratio of incremental customers to total deal customers.
• Deriving the BISR condition is relatively straightforward.  Executives desire the fraction of campaign profits associated with incremental sales to exceed the campaign costs.  Rearranging this conditional reveals that the BISR is the ratio of the campaign costs to campaign profits.
• If the fraction of deal customers that are incremental is above the BISR and the ROI is positive, then the firm can conclude that the campaign enhanced profits.  Otherwise, the campaign harmed profits.
• For Spring Restaurant’s Groupon campaign, we calculate the BISR at 75%.  Similarly, for the hypothetical newspaper coupon campaign, we calculate BISR at 74%.  (See cells C41 and E41.)  Thus, for either of these couponing approaches, more than 3 out of 4 deal customers must be incremental, or customers that wouldn’t have purchased in the absence of the coupon, in order for the campaign to be declared a good deal.

Groupon is Radically Different from Traditional Coupon Distribution, Yet Also the Same

The above models for analyzing couponing effectiveness highlight the differences between Groupon and traditional couponing, yet also demonstrate how similar the two approaches are.

Groupon differs strongly from traditional couponing in redemption rates and distribution costs.

• As noted, traditional coupons have a 1% redemption rate, while Groupon redemption rates are believed to exceed 70%.  This speaks both to the effectiveness of Groupon in targeting prospective customers for products, be they commonly purchased goods or unique experience oriented services.  It also speaks to the consumer behavioral differences between customers which purchase a Groupon from those which receive “free” coupons through traditional means.  Groupon is clearly an efficient means to attract interested customers.
• With traditional coupons, firms must pay for their distribution.  With Groupon, firms are paid for the rights to distribute coupons. This dramatically lowers the perceived risk associated with a Groupon campaign over that with a traditional coupon campaign.  This lower perceived risk for the couponing firm may account for the explosive growth of Groupon while also indicating the potential for this growth to be no more than a bubble.  Groupons still hold risk for the firm.  If the Groupon is not properly structured or if a Groupon is distributed for the wrong type of offering, it can be very damaging to profits rather than helpful to profits, as we will see.
• Furthermore, the fact that Groupon pays to distribute coupons for the firm implies that a couponing firm can make money from Groupon even if no one redeems the coupon.  This fact implies a potential area of tension where firms would like to have restrictions on their Groupon in order to lower redemption rates, while Groupon would like to see high redemption rates in order to encourage positive consumer feelings towards their Groupon experience.  These tensions are currently being played out in the market place and the courts, such as the recent lawsuit filed in Connecticut.

Groupon is similar to traditional coupon in terms of their purpose and approach to analysis.

• Groupons, like traditional coupons, are designed to induce trial that leads to repeat business.  The more new customers that Groupon creates that are like the other customers which have visited the establishment, the more likely the Groupon campaign will contribute to the success (profit) of the firm.  If Groupon customers are customers that would have purchased anyway, or if they are otherwise dissimilar to standard customers (lower average bill, less likely to return as full paying customers, etc), then the Groupon campaign is likely to be damaging (loss) to the firm.
• In terms of modeling the effectiveness of a promotional campaign, Groupon and traditional coupon campaigns can be evaluated on a similar basis.  Cost per buyer, ROI, and BISR are all calculated in a relatively similar manner.  Therefore, managers can and should compare Groupon campaigns to similarly designed traditional coupon campaigns in choosing between promotional activities.

Groupon Can Be Terrific, and Can Be Destructive

The analysis presented for Spring Restaurant demonstrated relatively equal outcomes for running a Groupon campaign versus a traditional coupon campaign.  (Outside of a lower volume of new customers associated with a single run newspaper coupon, the cost per buyer, ROI, and BISR were relatively the same.  To drive higher volume with a traditional newspaper coupon and bring it in-line with their Groupon results, Spring Restaurant would only need to run the advertisement more often.)  This is largely a result of the parameters used for conducting the analysis.

If we assume different input parameters, we find that Groupon can be much more profitable than a newspaper campaign, or that it can be very damaging to profits in comparison to a newspaper campaign.  (Readers are encouraged to change the parameters in the spreadsheet model to demonstrate the following scenarios.)

For instance, consider a traditional newspaper coupon campaign that had a very low redemption rate.  If Spring Restaurant newspaper coupon generated 30 or fewer customers redeem the coupon (a 4 in 10,000 redemption rate), then the newspaper coupon effort would have been a loss.  Firms like Spring Restaurant are likely to have very low redemption rates from traditional coupon distribution simply because their target market is sparse.  In comparison to consumer packaged goods, the demand for fine dining, skydiving, and other unique services is low.  Not everyone would enjoy a fine dining experience at Spring Restaurant, but most everyone does need soap.  Hence, the demand for Groupon and similar services is likely to remain dominant among high-margin, rarely purchased, service establishments in comparison to mass product firms (such as the Gap or L’Oreal).

Alternatively, consider a Groupon promotion for a lower-value restaurant, such as casual dining at Chili’s or fast food at the Chicago Bagel Authority.  A Groupon for a $20 dining experience sold for $10 to customers with similar restaurant gross margins and redemption rates would actually induce a loss.  In contrast a $10 coupon distributed via the newspaper with a 1% redemption rate is potentially profitable.  Hence, the demand for Groupon and similar services is likely to remain dominant among high-value establishments rather than low-value establishments.

Organizational Strategy for Promotional Effectiveness Analysis

The model presented clarifies the value of coupon promotions.  With a few modifications to fit a specific firm’s situation and expectations, it can be used to evaluate promotional effectiveness and identify good promotions.

As a gut reaction, managers are always looking for new ways of driving revenue and promotions, especially hot new forms of promotion such as Groupon, appear attractive.  This analysis doesn’t destroy the attractiveness of Groupon, it only informs it.  The model enables managers to demonstrate when a Groupon is likely to be effective and when it is not.  Hence, it informs our gut reaction to the proposition of couponing.

As a process, firms should use this approach to analyzing the effectiveness of all coupon promotions, even if they aren’t using Groupon.  (Groupon is just the example.  The approach can be used for most any form of price promotions with a few modifications.)  For instance, consider the following organizational strategy:

• A marketing manager drafts a coupon campaign plan including anticipated distribution costs, coupon face value, redemption rates, and expected percentage of coupon redemptions associated with incremental sales versus deal buying by regular customers.
• A pricing analyst would then use these numbers along with information on product prices and margins to generate the ROI and BISR.
• Campaigns which are anticipated to deliver a positive ROI and exceed the BISR hurdle would be given the green light.  Those that don’t would be sent back for rethinking.

To implement the above strategy, firms may have to implement two efforts:

• Establish a pricing center of excellence, or at least assign an individual, to run the analysis, track performance, share expectations, and guide future promotion decisions for marketing managers.
• Conduct a market research effort to determine the baseline expectations on the fraction of coupon redeemers that are incremental versus those that are not.

These organizational changes are being made by leading companies today.  With this approach, managers can confidently determine if Groupon is better than traditional couponing, or not.  They show the money to be made or lost in couponing.

Download the spreadsheet Wiglaf Pricing Explores Groupon v Newspaper

Note of Interest and Holdings: At the time of writing, the author is not currently a direct consultant to nor investor in any of the firms listed in this article.