RSR Research: Personalization vs. Relevancy
Through
a special arrangement, presented here for discussion is a summary of an article
from Retail Paradox, Retail Systems Research’s weekly analysis
on emerging issues facing retailers.
I have been sorely tempted to declare this
year the "Year of Retail Relevancy." Retailers
are getting a better picture of customer behavior in their stores, thanks to
more technology touchpoints that can capture what consumers are up to (including
consumers’ personal devices). Retailers have more data about consumers
than ever before, and are (theoretically) getting smarter about using it well.
Retailers also understand that while consumers have concerns about privacy and
how their data is getting used, they also desire for their favorite retailers
to demonstrate that they "know" them.
This last concept, at the surface
level, seems contradictory. Actually, I’ve
had a lot of complaints from retailers about this contradiction. The conversation
goes something like this:
Me: All of our research shows that retail winners are working hard
to provide more relevant experiences to individual customers.
Retailer: Yeah, but all our customers tell us that they don’t
want us to violate their privacy. So how do we create personalized experiences
without getting personal?
Well, that hits on exactly the problem. Personalization
does not equate to relevancy. However, done well, relevancy can feel like a
personalized experience to a shopper.
Let me give you a classic example. My
household is a die-hard Coca-Cola household. We have never once, in the 15
years that I have been married, purchased anything other than Coke. So as a
retailer, you might have Pepsi breathing down your neck (and paying you fistfuls
of cash) to give me an offer that might tempt me to switch from Coke to Pepsi
— to try a new Pepsi line or flavor, something, anything to try to win me
away from Coke.
If you approach me through the lens of personalization based on purchase
history, but no context to that purchase history, you might be tempted to take
the money and deliver me the offer. I may not redeem it, but it doesn’t
hurt, right? And if I do redeem it, it’s no skin off the retailer’s
back because you sell a product either way. Wrong. And that’s why this
is the year of relevancy in retail.
If I have been your customer for, let’s
say, six years, and in that time I have never once purchased a Pepsi product,
what is my perception of you when you hand me that offer for Pepsi? Increasingly
it is, "I have given you
six years of my grocery business, week in and week out, and you have just given
me an offer you know I will never use. Don’t you know anything about
me?"
That’s the difference between personalization and relevancy.
If you
approach your customers with only the idea of selling them more stuff, then
you’ve already lost the battle. That’s the narrow
vision of personalization: how can I use their purchase history to sell them
more stuff? If instead you approach your customers with the intention of helping
them out, whether that yields an immediate sale or not, then you are approaching
them through the much broader vision of relevancy. And customers will notice.
And keep coming back for more.
Discussion Questions
Discussion Questions: How can retailers create more personalized experiences without getting too personal? What do you think of the overall goal of ’relevancy’
for retailers?
Count me in as a skeptic when it comes to highly personalized marketing efforts. Everyone should carefully read and understand the work of Andrew Ehrenberg (see article in yesterday’s Ad Age: ) before plunging into such a program. All firms (regardless of what industry or geography they operate in) are much better off creating a big broad footprint for themselves (i.e., maximizing penetration) before engaging in a unique dialogue with each and every customer.
This is an interesting topic and goes against what many have been doing for years. Coupons at the register for a competing product to the one you just purchased. I agree retailers should be more focused on understanding each customers purchase behavior and then support that behavior. In the Coke loyal customer example, the retailer might instead work with Coke to offer the customer free store brand bread as a thank you for being a loyal Retail and Coke customer.
Panera Bread has a relatively new loyalty card program that I think all retailers should take a serious look at. They don’t email or snail mail coupons to me or even hand me a coupon for a future visit. They simply say “surprise,” the coffee is on us today. Or “surprise” you have a free bagel that you can claim during the next month. “Don’t worry about saving anything we will remind you each visit that the free Bagel is waiting for you.”
Imagine if I was checking out at my local grocery store and the cashier said “Surprise” you bought our Private Label Bread and it Free today as a thank you. Or “Surprise” we know you like Heinz Ketchup so we have a free one waiting for you when you are ready during the next 30 days. No coupons, no emails, no snail mail. Just random surprises when you least expect them.
Does it work? Let’s put it this way: When I am traveling or even in my local town, you bet I drop into Panera Bread for my coffee and hopefully a “surprise.”
Retailers can create a more personalized experience by explaining why they gather customer information and how it how it will be used. Then they need to deliver on that promise by providing relevant offers that have meaning and value to individual consumers.
What a wonderful anecdote about retail relevance! It illustrates the requirement to tag customers as loyals at multiple levels. Traditional CRM applications identify their loyal customers based on a high order shopping patterns. Do they shop often, recently, and greater that a certain volume of business? This is RFM analysis. But RFM does not get you to relevancy.
To achieve relevancy. the analysis of market basket must delve deeper. Is the customer loyal to certain brands as in Nikki’s story about her family’s affinity to Coca Cola? Does the shopper disproportionately respond to offers? Does the purchase history indicate a pattern of buying “solutions” in contrast to SKUs? Outfits or items? Kits or parts? Is the shopper a gift buyer? Is the shopper loyal to a particular salesperson and/or store? Does the customer shop both online and in store?
A retailer striving for relevancy would be wise to tag the customer records with much greater discernment than classical RFM.
Brava, Nikki, for this refreshingly clear discussion of a crucial issue. Segmentation and personalization are relatively easy, while true relevancy requires application of principles that remain mysterious to many marketing pros.
The interrelationship with privacy will require some further discussion, IMO.
Very fine article today here on RetailWire about Personalization vs. Relevancy. What we have found at David Biernbaum & Associates is that consumers absolutely embrace retailers that know them, know what they like, and know what they want to buy, albeit with a “catch.” Consumers don’t want to believe that they are being spied on, scrutinized, or that their privacy is being invaded. So retailers have to be smart enough to know their consumers without offending consumers. The way to do it is simple: explain to them the benefits (for them) to convey the information, and then ask them for it!
This is a great piece, a great question, and a common pitfall with personalization. How can retailers do more helping and less hawking, especially with regard to offers?
Shoppers say they they most value offers on items they regularly buy, offers related to personal events, and recommendations on relevant new items, according to a recent IBM survey of 30K shoppers (Institute for Business Value 2011).
How can you predict the best offers? I believe there are generally 4 ways:
1. Present a lot of offers, adjust a priori by season, segment, etc. and let shoppers sort them out.
2. Let shoppers opt-in to the kinds of offers they wish to receive.
3. Use shopper purchase history to recommend the most relevant offers.
4. Use collaborative filtering or similar techniques to suggest what like shoppers have chosen.
Options 2, 3, or 4 have all been called “personalization.” They work best depending on the information availability, privacy sensitivity, and target shoppers’ willingness to receive recommendations.
But the best approaches do something else. They intervene to tune out annoyances. This includes share-of-requirements filters that kill Pepsi offers to Coke enthusiasts, recurrence filters that eliminate repetitive offers, and the like. It takes both the right experience and intentions–the one-to-one marketers who listen to shoppers and anticipate ways to help rather than just hawk.
The key to “knowing” your customers without making them feel uneasy is to stick to purchase history. Most people don’t mind being sent personalized offers and suggestions based on what they’ve bought (as Nikki said, trying to tempt people to switch from a favored brand may have negative consequences). When people start getting targeted based on ethnicity, age, number of kids they have, etc. is when they start getting creeped out. Personalization based on where a store is located is also OK, but targeting people based on their home address may also approach the creep-out zone.
I completely agree that relevance is vital, and I want to reinforce that relevance needs to be achieved across the entire customer experience and sustained over time. Random acts of relevance where one part of the business is doing it well but another isn’t will still leave customers baffled and frustrated.
Retailers are only as strong as their weakest link, so making sure that the in-store and in-home experiences are both relevant is critical. This dictates that the business needs to be very well connected around the customer: a challenge that is notoriously hard in a product/category and functionally organized business. This requires not only a new approach to segmentation, insight and targeting, but an organization evolution with changes to KPIs and incentives that measure relevance to loyal customers.
The challenge of being relevant is only getting harder with all the new channels of communication available and with increasing diversification of their brands into credit, warranties and services. While the use of personalization is necessary in some of these areas, it can’t be abused in others.
Looking at the business through the lens of the customer is always a great place to start!
I’d argue most products are relevant to most shoppers; the question the savvy marketer must ask is “HOW is my product relevant to THIS shopper?”
For our clients, we advocate what we call a “portfolio approach.” Relative to a particular product, you can often divide the world of shoppers into a number of different groups, and then craft a different marketing message or offer for each group:
1) Heavy/Loyal current buyers. Give these shoppers a small reward, along with a simple “Thank you for being a great customer.” These offers are most strongly perceived by shoppers as being personalized. And, contrary to many marketers’ instincts (“I don’t want to waste money marketing to people who already buy my products!”), we find these offers regularly generate strong incremental sales.
2) Light current buyers. Try to up-sell these shoppers to a larger size or to a multi-unit purchase. Again, these offers are generally perceived to be personalized, and get good response.
3) Former buyers who’ve stopped. A moderate value offer often re-engages these shoppers, and yields strong incremental sales.
4) Category/competitive buyers. Hitting shoppers who buy a competing item (e.g., the Pepsi offer to Nikki), must be done carefully. Think of it as a sampling effort, and give the highest possible discount. Nikki might bristle (and rightly so) at a “$0.25 off a 12-pack” coupon from Pepsi, but an offer that says “We’re so sure you’ll love Pepsi that we want you to try a 6-pack, FREE!” might at least get credit for trying. To Nikki’s core point, though, these types of offers tend to have lower response rates.
5) Non-category buyers. When trying to engage shoppers who don’t even shop the category, a marketer should look for shoppers who buy similar items in other parts of the store. Keying off qualities of the product (Kosher, low-salt, vegetarian, organic, etc.) and delivering appropriately-crafted offers to shoppers with each preference can often yield surprisingly positive results.
There are more, but as you can see here, even a simple sub-division of the shopping population lets you be more personalized and relevant without sacrificing reach.
Nikki, always fun and engaging to read your posts. I could not agree more. Our work at Geomentum focuses on the intersection of the store and the neighborhood rather than the individual as the starting point precisely so that we can deliver relevancy without violating privacy. No one likes to be stalked and to your point about Coke, no one likes to feel their loyalty is taken for granted either.
Step one is about getting the right mix of product, media and marketing in the store depending on the kind of neighborhood it’s in. Once someone opts in or signals their interest via a search, a mobile coupon, a social engagement then you have the opportunity to get more personal but for many retailers the bigger win (and better margin play) is to forget about personalization and focus on getting me the products, services and messages most likely to matter to me right now (sounds obvious but if you look at the spring gardening stuff on display right now in Chicago you realize that local retailing is still an oxymoron for the national chains).
Great discussion–keep it coming!
Nikki Baird’s first sentence of the last paragraph says it all. Until retailers of all types learn that it is about SERVING the customer not pushing more stuff on them they are doomed to repeat the same mistakes, get the same unsatisfying results, experience the same frustrations, and wander the same wilderness of the clueless.
In addition to what has already been mentioned, relevance can be maintained for a shopper who has a particular brand preference by offering promotions of competing companies, but not necessarily the competing individual product. What about another PepsiCo Brand, like Gatorade, Frito Lay, etc.?
Every attempt to “personalize” or increase “relevancy” suffers some degree (small or large) of failure merely because it has to follow rules. Those rules may be in a computer. They may be in the agreement of a store with its suppliers. They may be merely rules of behavior for store staff.
But the beautiful human truth is that we’re much more varied than any of those lists of rules can fully comprehend.
This doesn’t mean we should ignore personalizing nor improve our personalizing with an eye to relevancy. But it does mean that we should carefully choose when and how we use those efforts.
Too much tech activity has been sold with the myth that computers can help us perfectly understand buyers/web site visitors/anyone online and therefore present them with perfectly relevant material. It hasn’t happened. And there’s no evidence it will.
As one example of “not happened,” look at Facebook’s miserable .05% clickthrough rate. These are supposed to highly targeted ads based on some of the best and most personal consumer information available. But I have TV campaigns that generates SALES from .5% of impressions.
Seems we need a new humility to recognize that all these perfect personalization/relevancy promises simply aren’t going to come through. Assuming that, we should adjust our expectations.
Chances are there are retailers out there you have a “sticky” relationship with and others you don’t. You and I generally choose who we want to be close to. There are some retail stores and brands I love, particularly those that serve my lifestyle and hobbies. Those are the brands I want to have a “sticky” relationship with. It is really up to the consumer to tell you whether they want this sort of relationship with your brand. And my use of the word “relationship” is not coincidental either. Before you can get more “personal” with somebody, you do need to develop a “consensual” relationship.
I am hoping you get my meaning. Not every customer is going to be responsive to your offer to be personal, to have a “sticky” relationship with you. Those who do, will be good customers and solid brand ambassadors. Embrace customers who consent, respect the wishes of those who don’t. It’s not only the right thing to do, it’s also good business.
This is a great article and touches on one of my favorite topics–technology vs. privacy. Increasingly we are able to do things from a technology perspective that may not be “right” from a privacy perspective. How do we strike the balance and create a solution that provides a relevant experience that does not encroach on consumer privacy? The FTC has recently published guidelines for online behavioral advertising, (which make for an interesting read) but hasn’t translated that to face-to-face transactions. It will be interesting to watch the evolution of personalization in the shopping experience juxtaposed against the evolution of increasingly stringent privacy requirements.
As developer of a personalized local deals site with the sole purpose of delivering relevancy to the user, I completely agree with this, and I’d like to elaborate. Consumers love personalization when translates to relevance and the younger demographics are beginning to expect this type of experience. The benefit for retailers is that if their messages are not competing with lots of irrelevant ones, they get more attention from the consumer.
Great issue Nikki.
My feeling is that shoppers don’t want their privacy “violated” but they do want the retailer to “know” them. Consumers are clearly willing to share personal info, the magic question is what you do with it.
1. If shoppers give you personal info, and then don’t get any benefit that’s a violation. If I give you my phone number or e-mail address, and then have to answer questions you should already know, then I feel I wasted my time trusting you with my info.
2. If shoppers perceive their privacy isn’t being protected (SPAM, Fraud, etc..) that’s a HUGE violation.
I think it was frog design that wrote about it becoming more common to say that people “trade” or “spend” information about themselves in exchange for online services. That metaphor implies that people are actually aware that an exchange is taking place. The key is that it has to be a “fair” exchange in the mind of the shopper.