Retention and Relationships

I’m starting a fairly long series of tips for retention and loyalty program managers on the corporate blog over at Loyalty Lab.  The first discussed program fund rates and is a good starting point for someone thinking about putting a program together.

With a day to think about it, I underemphasized the impact of program design on fund rate.  The art of design in this case is finding a balance between perceived fund rate (what the customer sees as the fund rate) with the actual fund rate (the resultant rate after breakage) while still meeting program objectives and staying fair to the customer.

A lot has to do with whether a redemption is a good thing or a bad thing.

In retail, you have to assess how much incremental revenue a redemption will generate, either through incremental spend during the visit (i.e. customers typically spend $50 when redeeming a $5 reward) or through incremental visits (i.e. overall member frequency increases by 0.75 visits/year when they reach a minimum spending threshold).  For many retailers, generating visits is so expensive that they want redemption, since that generates a visit.  In this case they want a low breakage rate.

In airlines, a redemption is pure cost, albeit low (the incremental cost of filling another seat is very low, so as long as they leverage existing capacity effectively, costs are not high).  The main benefit to the airline is share of travel wallet, so there is little benefit on the back end other than keeping customers happy and maintaining positive brand image in the social sphere.  Breakage here is a good thing, since it keeps costs down.

Other industries have similar considerations.

I’ve always visualized this as a multi-dimensional optimization problem, meaning the real challenge is framing the problem properly.  That’s why we’ve always spent a good amount of time understanding the behaviors that matter, the revenue generation possible, and the sources of cost in order to drive towards a set of solution options to consider.

I listened to an Invisible Hand podcast today with Douglas Hubbard on How to Measure Anything, and was struck by the similarities between his methodology and how we have to approach modeling of client businesses in relation to customer loyalty.  Topic for another day…

If there’s one unexpected mental shortcut I’ve learned from examining dozens of corporate loyalty efforts and customer databases, its that just about any idea will work for 20% of customers. The problem for marketers? You can’t predict which 20% with a lot of certainty, and there’s only partial overlap between any particular 20%.

Price promotion? 20% (probably more, but you get the point)
Loyalty program? 20%
Recognition? 20%
Special event? 20%
Early access to new releases? 20%

The list goes on. A few get more, many get less, but 20% is a good rule of thumb. Why 20%? Hard to say, but that’s what the data says.

The takeaways?
1) Incorporate as many interesting ideas as you can manage effectively, since each will produce results. But don’t stretch too thin, or all of them will drop off.
2) Choose ideas that do not overlap, to ensure maximum return and responsiveness.
3) Budget has an impact, no doubt. Stack ranking from cheapest to most expensive isn’t a bad idea.

There’s a whole separate group of initiatives that only 1% will respond to…that’s for a later post.

My latest at VentureBeat’s Entrepreneur Corner.

This was an interesting article, primarily since it was very focused on a specific task – generating a second sale.  While repeat visits often happen naturally, nudging customers to return is often necessary.  The second purchase is incredibly important, since it identifies the customer as someone with the potential for a long term relationship.  Additional investment in the relationship will have a far higher likelihood of paying off with these customers.

[Things are quite busy on the work front…but I will get back on schedule starting with this post.]

Direct communication, possible once a customer is identified, unlocks a whole new level of interaction, where marketers can differentiate the message.  The various options now available enable a much more nuanced relationship with your customers.  Looking at these options is instructive in understanding the value that this element can add to your overall best customer management.

Triggered email is the high ROI workhorse of differentiated communications.  Expectations for email relevance will only continue to grow, so marketers need to continuously increase their capability to develop and deliver messages that reflect actions, attributes, and environmental factors.

Custom printed or variable printed direct mail is a personalized variation on traditional direct mail.  Ideally the same factors that make triggered email relevant are incorporated into direct mail content, either through differentiated messaging and creative, or “statement”-type data for a formal program.  In reality direct mail can and should eventually have the same level of flexibility as email.  Good marketers will incorporate this capability into their arsenal for one good reason – it works.

Targeted promotions are an evolved version of coupons, with the qualification and award tuned to the individual recipient.  While most marketers can only provide relevance through the targeting (that is, identify segments and craft promotions for each), we’re not far from algorithmically generated promotions taking over.

Site personalization is the one area that should be far more prevalent than it is.  Variable site content has been around for a long time, yet only a few companies capitalize on the increases in conversion and repeat visitation that can result.  If everything is going right, the site is reflecting the activity going on elsewhere, such as the targeted promotions above, with landing pages for the individualized triggered email.

There are other types of direct communication that are still emerging.  Widgets, small client-side apps that can be added to websites directly or via some basic integration, are quickly gaining traction and critical mass.  While many are just distributed links to a site home page, more interesting technologies are available when a company opens up their APIs, with Facebook as the current shining example.

Knowing which individuals are your customers is a crucial task for every company.  It’s easy to do in business-to-business companies, but can be quite difficult for consumer-focused companies. 

There are two parts to this problem – who is the customer, and what did they buy?  While you can solve the first part of the problem without the second, that’s not recommended.  You will have no way to calculate customer value, making future optimization decisions much more difficult.

So the best approach is to tackle both parts simultaneously.  In retail, loyalty programs and credit cards were the traditional way to gather customer identities and connect them to individual transactions.  This approach still works.  The other major approach, data appends, now depends mostly on reverse appending phone numbers, which is much tougher due to rising cell phone penetration (which can rarely be appended).  In certain circumstances, email can be used as the tracking mechanism outside of a loyalty program.

Companies can reduce the barriers to identification with newer technology (such as new point of sale systems) which can be expensive, or with great relationship programs and a great brand.  The former increases the capture rate at the point of purchase, the latter increases a customer’s willingness to raise their hand and be tracked voluntarily.

The mix that makes the most sense for your business depends on your current capabilities, budget, and overall objectives.

If you have nothing today, I’d start with a relationship program with soft benefits, requiring an email address.  If you have a high-throughput point of sale/interaction, think about using phone number as another identifier.  Bar coded cards are the most accurate, but not all consumers are willing to carry them, so you will need a ubiquitous backup identifier, and phone number has the benefit of being reverse appendable (no, its not a word, but it should be).  So at minimum, you could connect multiple purchases over time to a specific phone number, even if you can’t identify the individual.

Down the road, you will end up with more records than customers.  At that point, bring in a professional data hygiene firm to help with merge purge.

My latest at Chief Marketer.  Its the third in the series on developing a loyalty strategy proposal.  The three taken together provide a framework for addressing all of the key questions regarding a loyalty (or relationship marketing) program.

At its core, ROI isn’t that tough to figure out, but the secret sauce is estimating the impact on transaction size, frequency, and retention.  There’s nowhere you can find those numbers online, since every company has a unique mix of factors that influence the values.  Experienced practitioners can do that estimation for you.  If they don’t ask you what your plans are for employee training as part of the program, then be skeptical about what they tell you…

My latest at Chief Marketer.  Part two of a three column series that walks through a basic approach to loyalty program strategy.  This one looks at the different program structure options and discusses how they are applied based on the segment-specific strategies developed in the first column.

I’ve already received several requests for clarification on the customer grid from my Chief Marketer column.  Once you have chosen two criteria for segmenting customers, imagine a grid with each criterion on one axis.  Each combination of criterion values represents a segment to be investigated.

So if a hypothetical sporting goods store was set up where Criterion A is Private Label Credit Card Holder = {Y,N} and Criterion B is Activity Type = {Outdoorsman, Weekend Warrior, Family Sports}, there would be 6 boxes in the grid.  (Resemblance to a company where I once ran marketing is completely coincidence.  Besides, that company no longer exists.)  For each box, provide an objective and develop the strategy to accomplish each objective. 

For example, the Outdoorsman PLCC Holder has a lot of Open To Spend and lots of big ticket items in his typical wish list.  So the objective might be to generate big ticket purchases.  The loyalty strategy here might be to focus on long-term accrual, so that he can save up benefits to use on his one big purchase.  Or you might assume he has already made at least one big ticket purchase (a possible third criterion for your grid) and your objective is to maintain his spending level.  The loyalty strategy here might be to provide point acceleration on key consumables or accessories throughout the outdoor departments or provide increased accrual rates when a spending threshold is met.

My most recent column at Chief Marketer.

I’m finding more and more of our new clients and current prospects would benefit from additional insights into basic loyalty strategy, which prompted this most recent column.

My latest article at Chief Marketer.

The gist of it is a series of non-obvious benefits to loyalty marketing:

Loyalty programs can be very cost-effective, especially after breakage.

Loyalty programs get you more e-mail addresses of buyers.

Points programs let you promote without actually lowering prices.

Loyalty programs build lock-in, especially important for easily substituted businesses.

Loyalty programs improve the customer tracking rate.

Loyalty programs make it easier to reward nonpurchasing—but valued–behavior.

Loyalty programs let you give status to your best customers.

Loyalty programs greatly increase the likelihood of a message being read.

Surprise and delights appeal on a very deep level.

Loyalty programs tend to attract those who are most likely to respond to them.

The details of each of these points are discussed at Chief Marketer.  Enjoy.