Category Archives: CXM

Why I’m Advising Bluecore

I first read The One to One Future by Don Peppers and Martha Rogers in 1997, four years after it was published.  As a marketer, the book made a big impression on me.  It was revolutionary stuff:  we should make the paradigm shift from mass marketing to individualized marketing.

When the book was published in 1993, newspaper ads were $75B/year, TV around $60B, the web browser was a mere three years old, and there were 623 total sites on the web.  There was effectively no web advertising market.  It was nine years before the Minority Report popularized a future vision of one-to-one advertising.  It was six years before Paco Underhill published Why We Buy revealing insights gleaned by manually tracking shoppers to understand in-store behavior [1].

Look at the subtitle: “Building Relationships One Customer at a Time.” You could use that in a webinar today.  The One to One Future was not just ahead of its time; it was so far ahead of its time that it could have equally been categorized under either “marketing” or “science fiction.”


  • It turns out, as with science fiction, that it’s easier to envision something than to build it. Remember, “they promised us flying cars and we got 140 characters.” [2]
  • Building individualized marketing systems required layers and layers of underpinnings that were simply not in place. You can’t do good personalization without a clean, real-time, 360-degree view of your customer.  Clean means a big effort into data quality and data profiling and typically either master data management or a customer data platform [3].  Real-time means real-time data integration [4].  360-degrees means pulling relevant data from virtually all of your systems.  Self-driving cars don’t work on cow paths.  Building those layers of requisite infrastructure has taken decades.
  • Marketing’s focus on the perfect offer was flawed. Say I found an offer with an 90% chance that you’d respond affirmatively.  Perfect, right?  But it was for a product that was out of stock.  The perfect offer has to be for the right product, in the customer’s preferred size or color, and available to sell.  We can’t just find the set called {great offers}.  We needed to intersect it with the set called {in stock and need to sell}.  This made a hard problem harder by pulling inventory and the supply chain into the equation.
  • Marketers got trapped in a vicious downward cycle of communications. Email click rates have nearly been cut in half over the past decade.  Marketing’s solution?  Send more emails to make up the difference.  Email vendors, who typically price by the email, were only too happy to accommodate.  That, however, is a short-term mentality.  More bad email with lower open and click rates isn’t the solution.  The same holds for ads and promotions.  Marketing needs to get out of this race to the bottom.  We need to focus on quality, not quantity.  And pay vendors for performance delivered, not communications sent, while we’re at it.
  • Finally, the retail industry needed to shift mentality from store-first to digital-first. Roots, as they say, run deep and retailers have long, deep roots in physical stores.  Bricks-and-mortar supposedly changed to clicks-and-mortar, but really, it was mortar-and-clicks the whole time.  The industry never really changed to digital-first from store-first.  Until Covid-19, that is.  While this meme, popularized in Forbes, was intended for many industries, it could have been custom made for retail [5].

So where does Bluecore fit in?

  • Bluecore is a multi-channel personalization platform. They’re building what marketers in the past dreamed of, but couldn’t build, because the infrastructure wasn’t there.  Now it can be built, and they’re building it.
  • Bluecore is an AI/ML company focused on retail analytics and personalization. I’ve blogged before that AI/ML is best applied to specific problems and not general ones, and this is a great example.  They are a closed-loop, retailed-focused application that gets smarter every day and with each new customer.  If you believed in the increasing returns of marketing leadership in technology markets before AI/ML [6], you should believe in them twice as much after.
  • Bluecore’s personalization understands both customer and product – and intersects them. Across a catalog of more than 250M products and SKUs, Bluecore can match customers and products at a 1-1 level.  It automates what would have been the work of a team of in-house data scientists.
  • Bluecore is paid for performance, not volume. They back up their performance claims with a pricing model based not on volume but on success.  This is a great example of superior technology enabling disruptive business model innovation.

Why am I advising Bluecore?  Three reasons:

  • As a true, blue marketer this stuff genuinely interests me. I love working with marketing companies on marketing problems.
  • It’s always about the team. I’ve loved working with Fayez Mohamood (founder/CEO) and Sherene Hilal (SVP of Marketing).  As a bonus, former Salesforce teammate Scott Beechuk is an investor and on the board.  I like working with people who like working with me and appreciate my inimitable (I inadvertently almost typed inimical) style when it comes to feedback.
  • The momentum and market opportunity. Bluecore’s a highly successful company, having raised over $100M in VC with top-tier investors, and they are pursuing transformational change in a $4T market.  The last 100 years in retail were all about stores, the next 100 will be about retailers meeting customers wherever they are.  And that’s what Bluecore does.

# # #

[1] And why, to this day, you can find still baskets strewn throughout many retail shops as opposed to only at the entrance.  His work was kind of a manual predecessor to systems like RetailNext, whose founder I got to know through mutual investments in a prior life from StarVest.

[2]  Peter Thiel at Yale.

[3] Which weren’t to be invented for about 20 years

[4] The data warehouse was invented in 1992, with the publication of Bill Inmon’s Building the Data Warehouse.   Ralph Kimball would invent the star schema 4 years after that.

[5] Apologies to frequent readers for using this meme again – but I just love it!

[6] Tech buyers, and particularly IT buyers, tend to face high opportunity costs and high switching costs and are ergo generally risk averse.  This drives increasing returns for early market leaders.  Think:  no one ever got fired for buying IBM.

Hotwire’s Demonstration of How to Lose a Customer for Life

Here’s a simply superb recipe for how to lose a customer for life.

Step 1:  Misrepresent your offering.  

I needed to book a hotel on Hotwire.  I picked a 4-star blind booking because I didn’t care which hotel, I just wanted any nice hotel.  See the word “hotel” appearing twice on this screen.  They booked me into a condo. But I wanted a hotel — I’m not on Airbnb, I’m on Hotwire and I pressed a button called hotel.  I don’t want someone’s condo.  I want a hotel and what’s associated with it — e.g., 24-hour reception, a bar, a restaurant, lobby, etc.

I’m already unhappy and the story’s just beginning.


Step 2:  Fail to Deliver

I drove to the condo complex and my worst fears were realized.  There’s nothing hotel-ish about it.  It’s a series of residential towers.  No lobby, no restaurant, no bar, no life.  Heck, I had to struggle to find the check-in area.  I found an “agent on duty” sign but the check-in area was closed.  No worries, I called the number for the property listed on my reservation, twice, but got no answer.  I waited 15 minutes to see if anyone shows up.  No luck.  So I booked another hotel (not via Hotwire) and stayed there instead.

on dutycheck in closed

Step 3:  Repeatedly fail to handle the complaint

I emailed Hotwire requesting a refund because I was unable to access the property because the check-in area was closed and the property didn’t answer the phone.  Hotwire responded saying the property told them they have a 24-hour desk with security who will check people in.  I replied that was funny because I was there and I can tell you that they didn’t.  I mailed them pictures of the “agent on duty” sign and closed check-in area.  I reminded them that I called the number listed for the property twice and no one answered.

I also reminded them if they booked me into a hotel — as the screen said — and not a condo, then they would have had 24-hour reception and I’d never have had this problem in first place.  They refuse to do anything.

I don’t give up.  The world’s not perfect; I run a business — problems happen, it’s how you respond to them that really counts and I know we’re going to need to escalate beyond first-line support to get a refund.  I wrote again.  I asked my poor assistant to call them.  She was kept on hold for an hour, hung up on, call dropped.  Again, they refused to do anything.

Because the facts are so stark, I’m confident I’m going to win eventually and continue to use Hotwire during the ongoing dispute.

Step 4:  Issue this as final resolution

Eventually I get tired of the whole process and send the “last chance to fix this or I’m gone as a customer” letter.  Here is Hotwire’s rather amazing reply:

hotwire lame email2

Let’s review this for a second:

  • You recognize that I was unable to stay at the property because no one was there to let me in.  This is not a satisfaction issue, this a a non-delivery of service issue.
  • You regret my frustration and you want to thank me for my feedback — I’m not giving feedback, I’m demanding a refund for breach of contract.
  • But “regardless of who you may speak with our response will remain unchanged.”   (I hope you cleared that with your CEO, because he or she might like the right to overrule you, but you’ve taken that away.)

Going forward, I’m not sure what I’ll do.  I think I’ll follow the ideas suggested in this post, which I wish I found a few months ago, such as complaining to the credit card company and reporting them to the Better Business Bureau.  But that’s not the point of this post, which is about customer experience.

Hotwire, think for a second.  What other response than “goodbye forever” could a customer possibly have to your email?

Bye.  Forever.

Please Give Me a 10:  Interpreting Customer Satisfaction Surveys in an Era of Bias

Say you’re considering going out to dinner in a city you’ve never visited before and there are two different surveys of local restaurants that you can use to help choose a place to eat:

  • Survey 1, which is taken by randomly asking customers leaving restaurants about their experience.
  • Survey 2, which was conducted by asking every restaurant to provide 25 customers to survey.

Which would you pick?  Survey 1, every time.  Right.  It’s obvious.

Why?  Because of what they measure:

  • Survey 1 measures customer satisfaction with the restaurant in an objective way and can be used to attempt to predict your experience if you eat there. In a perfect world, you could even slice the survey results by people-like-you (e.g., who liked the same restaurants or have similar food profiles) and then it would be an even-better predictor.
  • Survey 2 measures how well the restaurant can pick, prime, and potentially bribe (e.g., “three free meals if you take the survey and give us a 10”) its top customers. It has little predictive value.  It is more a measure of how well the restaurants play the survey game than the quality of the restaurant.

Would it surprise you to know that virtually every major survey in IT software is run like Survey 2?   From big-name analyst firms to respected boutiques, the vast majority of analysts run their customer satisfaction surveys like Survey 2.

Why would they do this, when it’s so obviously invalid?  Because it’s easier, particularly when you need to include a bunch of relatively small startups.  Finding a random list of Oracle and SAP customers isn’t that hard.  Try finding 20 customers of a startup that only has 50.  You can’t do it.

So you make do and ask vendors for a list of customers to survey.  You get a lot of data you can analyze and put into reports and/or awards.  More disturbingly, you can build your special two-by-two quadrant or concentric circle diagram leveraging the data your survey, lending it more legitimacy.  (Typically these diagrams have one more-objective and one more-subjective dimension and things like revenue/size/growth and customer satisfaction factor into the more-objective dimension.)

When people challenge your survey and the methodology behind it, you can typically defend yourself in one of several ways:

  • “The data is the data; I’ve got to work with what I have.” But the data is garbage because of the biased way in which it was collected and the first rule of data is you can’t analyze garbage – “garbage in, garbage out” as the maxim goes.
  • “It was a fair contest,” meaning that every vendor had the same opportunity to select, prime, coach, cajole, reward, and/or bribe the respondents.” While this may be an excellent stiff-arm for the vendors, end users don’t care if your survey was FAIR, they care if it is VALID – i.e., can it provide a reasonable prediction of their experience.  And, back to the vendors, are such contests even fair?  A low-end,vendor with 1000 small customers can cherry-pick its customer base more easily than an enterprise vendor with 75 big ones.
  • “The results are consistent with our general experience talking to customers.” This is a weak defense because it’s both subjective and certainly confirmation biased – what analyst wants to undermine his/her own survey?  It’s also problematic because the customers who call analysts are not random.  Some serve certain verticals or departments.  Some serve big IT groups.  An echo chamber is often created in that process.

In my opinion, the single best thing these surveys do is ferret out vendors that are marketing true vaporware (e.g., a mega-vendor with a new cloud product that they’ve given free to 300 customers in order to claim market success, but since no one is actually using it, they can’t even produce the 25 references).  For that these surveys work.  For everything else?  Not so much.

The whole situation reminds of buying a car where the dealership hits you mercilessly with:

  • “Is there anything I can do to make you more satisfied today?”
  • “Is there any reason you will not give me a 10 when corporate surveys you because my compensation will fall if I get anything other than a 10 and my lovely spouse and children (in the photo on my desk) will suffer greatly if that happens?
  • “You don’t want to hurt my children do you? So please give me a 10!”

Now I can guarantee you that the net promoter score (NPS) produced by that survey will not be valid.  So why do companies do it?  Because, like it or not, it does force a conversation where the dealership asks some important, uncomfortable questions that might highlight correctable problems.

If you’re trying to force a conversation between your organization and your customers, there is probably a role for the “please give me a 10” survey.  If, however, you are genuinely trying to measure satisfaction with your products, then there is not.

So what’s a buyer to do?  If you can’t trust these surveys, then what can you trust?  I think 3 things:

  • The vendor’s wheelhouse.  While most technology vendors attempt to position as everything to everyone, despite their misguided marketing they nevertheless develop a reputation for having a wheelhouse (i.e., an area or segment which is their real strength).  These reputations get built over time and are usually accurate, so ask people “what is vendor X’s wheelhouse?”  Not what do they say they do.  Not every area in which they have one customer — but their wheelhouse.  You should see a consistent pattern over time and you can then compare your needs to the vendor’s wheelhouse.
  • Reference customers. While I believe you can cajole someone into giving you a 10 on a survey, it’s much harder to cajole someone into bogus answers on a live reference call.  The key with reference checking is to find customers like you in terms of size, complexity, problem-solved, and general requirements.
  • Your own evaluation process. If you’ve run a good evaluation process, trust it.  Don’t let some survey up-end a process where you’ve determined that product X can solve problem Y after looking at demos, possibly do some sort of proof of concept, done vendor presentations and discovery sessions, built a statement of work, etc.

A Look at the Zendesk S-1 (IPO)

I thought I’d take a quick read of the Zendesk S-1 today, so here are my real-time notes on so doing.  Before diving in, let me provide a quick pointer to David Cummings’ summary of the same.

My notes:

  • 40,000 customers in 140 countries
  • 2012 revenues of $38.2M
  • 2013 revenues of $72.0M, 88% growth
  • 41% of revenues from international.  (High for a SaaS company at this size, but makes sense given their roots.)
  • Net loss of $24.4M and $22.6M in 2012 and 2013, -30% net loss in 2013
  • Zendesk approach:  beautifully simple, omni-channel, affordable, natively mobile, cloud-based, open, proactive, strategic.  They do this well.  (I’ve always viewed them as a very well run, low-end-up market entrant.)
  • Founded in Denmark in 2007.
  • 115M shares outstanding anticipated after the offering with seemingly another 40M in options under various options and ESOP plans.  (Seems like a lot of dilution looming.)
  • 65% gross margins.  (Though they don’t break out subscription vs. service which probably depresses things a tad.)
  • 20% of revenue spent on R&D.  (Normal.)
  • 52% of revenue on S&M.  (High, particularly for freemium which is notionally low-cost!)
  • 22% of revenue on G&A  (Normal to high, probably due to IPO itself.)
  • $53M in cash at 12/31/13
  • Headcount growth from 287 to 473 employees in year ended 12/31/13, up 68%
  • They have experienced security breaches:

“We have experienced significant breaches of our security measures and our customer service platform and live chat software are at risk for future breaches as a result of third-party action, employee, vendor, or contractor error, malfeasance, or other factors. For example, in February 2013, we experienced a security breach involving unauthorized access to three of our customers’ accounts and personal information of consumers maintained in those customer accounts.”

  • “[We are] highly dependent on free trials.”  (These guys define freemium model for enterprise software in my opinion.)
  • S&M org grew from 85 to 165 employees in period ending 12/31/13.
  • Owe $23.8M on a credit facility.  (Rare to see this much debt, but probably a smart way to reduce equity dilution.)
  • The three principles that drive the founders:  Have great products.  Care for your customers.  Attract a great team.  (Beats “Don’t Be Evil” any day in my book.)
  • Dollar-based “net expansion rate” (closest thing they discuss relative to renewals or churn):

    “We calculate our dollar-based net expansion rate by dividing our retained revenue net of contraction and churn by our base revenue. We define our base revenue as the aggregate monthly recurring revenue of our customer base as of the date one year prior to the date of calculation. We define our retained revenue net of contraction and churn as the aggregate monthly recurring revenue of the same customer base included in our measure of base revenue at the end of the annual period being measured. Our dollar-based net expansion rate is also adjusted to eliminate the effect of certain activities that we identify involving the transfer of agents between customer accounts, consolidation of customer accounts, or the split of a single customer account into multiple customer accounts. […] Our dollar-based net expansion rate was 126% and 123% as of December 31, 2012 and 2013, respectively. We expect our dollar-based net expansion rate to decline over time as our aggregate monthly recurring revenue grows.”

  • $66M accumulated deficit
  • Have data centers in North America, Europe, and Asia
  • 4Q13/4Q12 growth rate = 83% compared to 2013/2012 growth rate = 88%.  (Suggests growth is gently decelerating.)
  • Cashflow from operations in 2013 = $4.0M.
  • But they had -$24.1M in cashflow from investing activities.  (This is confusing because it’s a mix of items but broken into $12.4M in “marketable securities, property and equipment,” $7.1M to build data centers, and $4.7M in capitalized software development.  I’m not an accountant but if you ask me if “the business” is cashflow positive, the answer is no despite the $4.0M positive cashflow from operations. Building data centers and developing software, regardless of accounting classification, are all part of running the business to me.)
  • I am surprised they capitalize R&D.  Most software companies, far as I know, don’t.

zendesk common fmv


The FMV of the common stock is depicted above, by my math an annual 68% appreciation rate.

  • Huge number of leads are organic:  “the quarter ended December 31, 2013, 70% of our qualified sales leads, which are largely comprised of prospects that commence a free trial of our customer service platform, came from organic search, customer referrals, and other unpaid sources.”
  • SVPs listed (CFO, R&D) earn $240K base + $40K bonus
  • Automatic 5% share expansion / “overhang” built into the stock option and incentive plan.  Pretty rich in my experience and haven’t noticed anyone else doing it automatically before.
  • Letting execs buy stock with promissory notes … hum, I thought that went out with leg warmers.  Both loans were paid off by 12/31/31 and maybe that’s why.
  • CEO will own 7.1% of shares after the offering, including 4.3M (of the 8.1M beneficially owned) granted as options at the 2/14 board meeting.  (Seems odd to me; a huge option grant right before the IPO.  Hum.)
  • Nice banker line-up:  Goldman Sachs, Morgan Stanley, Credit Suisse, Pacific Crest
  • Raised $71M in preferred equity / venture capital
  • They do monthly, quarterly, and annual invoicing.  (Surprised they offer the short terms, particularly monthly.)
  • $6.5M in advertising expense in 2013
  • $11.4M in capitalized “internal use” software on the balance sheet at 12/31/13
  • They paid $16M for the Zopim (live chat) acquisition
  • Ticker symbol:  ZEN

Ten Customer Experience Lessons from My Recent Air France Nightmare

“This is how the customer relationship will end.  Not with a whimper, but a bang.”– adapted from TS Eliot

I’m not new to customer relationship management (CRM), customer loyalty à la Fred Reichheld, or customer experience management (CXM) as it is now more recently and holistically labeled.  I spent a lot of time working on CRM from the business intelligence (BI) angle during my near-decade running marketing at Business Objects.  Recently, I was general manager of customer service applications at

I’m also not new to France.  I’m an admitted Francophile, have lived in France for five years, and helped build one of — if not the — most successful enterprise software company in France.  I am thus quite familiar with French customer service where the problem is not that the experience is never good – some of the best customer experiences of my life have been in France — but rather that it’s highly bi-modal [1] and resists industrialization.

This post began life entitled “The Ten Reasons I Will Never Fly Air France Again.”  But the passage of time, the realization of a lack of alternatives on the SFO/CDG route, and a diving catch by the Air France Twitter Support Team, caused me to change it.  Nevertheless, it stems from a horrific experience I had this summer with our return flights from Corsica [2].  Instead of dwelling on the idiocy of the entire experience or its parts, I will attempt instead to extract lessons from it.   Details will be dumped to footnotes.

All customer interactions are moments of truth which combine to shapes the ultimate experience.  During very many of these moments of truth Air France failed and failed badly. Let us now try to learn from their errors.

Ten Customer Experience Lessons from My Recent Air France Nightmare

10.   Do not expose seams in partnerships or international operations

9.  Do not hide behind industry-speak to explain why you can’t fix my problem

8.   Do not tee-up failure with either inane or error-prone policies

7.   Do not value customers one-dimensionally

6.   Be ready to give when you are ready to take

5.  Don’t make your customer loyalty program the best reason NOT to use your service

4.   Admit fault where due and act accordingly

3.  Be mindful of my costs in dealing with you

2.  Empower your agents to fix problems, particularly supervisors

1.  Have a social media support team all with supervisor++ powers

Let’s now drill into each of these with more detail and specific examples, the most outrageous of which earn a #Kafka tag for the circular torture to which the customer is exposed.

Do Not Expose Seams in Partnerships or International Operations 

Air France showed major operational seams not only with their partner Delta but also with their own US operations.

  • When I call the (US) 800 number listed on your website, do not answer the phone “Delta/Air France.”  You know what number I called.  Answer the phone accordingly.
  • If you choose to use a partner (i.e., Delta) to take calls for you in a region, then empower them fully.    (After eventually getting a supervisor I said, “I presume you are more empowered than regular agents,” to which she replied, “Not on Air France; I can’t do anything.”)
  • Supervisors at your partner should have instant access and call-transfer ability to supervisors at Air France.  (At the end of our failed 90+ minute call to Air France that ended up in a Delta call center, the Supervisor put us on hold for 10 minutes “trying to find a number” we could call at Air France only to tell us she didn’t speak French and couldn’t find one on their website.)
  • Do not expose me to problems because the partner you chose to work with actually did something on my case.  (At one point, we were told we’d bought our tickets on Delta — which wasn’t true — and ergo that Air France couldn’t help us.  #Kafka)

Do Not Hide Behind Industry-Speak To Explain Why You Can’t Fix My Problem

As we suffered through this experience, I realized that most of the time we spent talking to call center agents was spent with giving us increasingly detailed and often different explanations – all buzz-speak laden – as to why they couldn’t fix our problem.

  • I don’t care who issued my ticket.
  • I don’t care if my ticket needs to be re-issued.
  • I don’t know what issuing a ticket means.
  • I don’t want to know what issuing a ticket means
  • Therefore I don’t care if the original agent making an issuing (or is it issuance) error is the root cause of the problem.
  • I don’t actually care what the root cause of the problem is.
  • I simply want you to tell me that I can fly on the flight on which I purchased my ticket and on which you had confirmed me multiple times.

Do Not Tee-Up Failure With Either Inane or Error-Prone Policies

Two inane policies were root causes of our troubles, far as I can tell.  The first was this kind of scheduling:

  • Schedule flight 99 at 2 PM
  • Subsequently reschedule flight 99 to 8 PM
  • And insert a new flight 97 at 2 PM

Please imagine the Who’s on First discussions this then drives between customers and agents.

“Hi, I’m on the 2pm flight on August 3rd.”

“That would be flight 97.”

“No, flight 99.”

“No, flight 99 is at 8 PM.”

“No, my paper says right here that flight 99 is at 2 PM.”

“No, flight 99 is at 8 PM … wait, when did you print your paper?”

“Right after I bought my ticket.”

“Well, I see your problem.  Back then flight 99 was at 2 PM.  But now it’s at 8 PM”

“But I booked the 2 PM because I wanted to travel at 2 PM, not because I wanted flight 99.”

“Well, the 2 PM is sold out.  I’m sorry.  You should have booked early.”

“But I booked six months ago and I was on it.”

“No you weren’t, you were on flight 99.”


By the way, even if you somehow thought it was a good idea to schedule flights in this way, wouldn’t it then also be a good idea to automatically move people from flight 99 to flight 97 – i.e., to tell them they had a flight number change as opposed to a departure time change?  (To see how Air France compounded this mess, read footnote [3].)

The second inane policy is an oldie-but-goodie.  Air France, has for decades at this point, had a “business” or premium economy seat within Europe that exactly the same as a regular economy seat.  No extra width [4].  No extra legroom.  No ability to advance-reserve seats.  And I believe (and certainly on our flights), no difference in food or beverage [5].

Air France needs to be called out on this:

Do Not Value Customers One Dimensionally

Airlines can get too wrapped up on miles / flight frequency as a measure of loyalty and value.  I believe yet another root cause for our troubles is that I am currently a low-frequency (i.e., bottom tier) frequent flyer on Air France.

Let’s do some math to illustrate my point.  Hypothetically, let’s say I make the decision and buy tickets for 6 people who travel to France every other year on $4500 discounted business class tickets.   That’s $27K every two years or $13.5K/year.  Let’s say Pierre travels for a small business, is careful about his fares, and flies 8x/year to SFO at an average of $1.5K/ticket.  Pierre is logging 88,000 miles/year on Air France and paying then $12K.  I am logging 2750 miles/year and paying $13.5K.  Pierre is a probably a penultimate tier flyer and treated pretty well.  I am treated like junk.  Yet, are we that different?  Who is, in fact, more valuable?

This is why all businesses need to be very careful about how they value customers and, to the extent they want to treat them differently as a function of value, not be one-dimensional in making such assessments.  Specifically, I think airlines [6] need to be more focused on:

  • Profit than revenue
  • Revenue than frequency
  • Decision makers than passengers

Because these can be hard to measure (e.g., decision maker) we tend to not measure them.  But I can assure you that in two years someone will be making the airline decision for the bi-annual France trip.  Just because identifying the decision maker is difficult, doesn’t mean that there isn’t one or that he/she doesn’t matter.

Be Ready To Give When You Are Ready To Take

This one is simple, yet broken with stunning frequency.  Here’s the rule:  if you’re ready to charge me for something when I take it, then you need to be equally ready to return my money if you take it back.

I mentioned this already in the context of Air France silently downgrading us from premium economy.  United did the same thing to my wife recently.  She is a bottom-tier United flyer, but had paid extra for a specific seat, paying something like $68 extra for a bulkhead only to find when she was about to board that someone else had been given her seat.  (I suspect because they were high-tier frequent flyer.)  They moved her back one row, without her consent, just handing her a new boarding pass.  But the seat they moved her to cost only an extra $30, and nary a mention was made of the $38 difference.  That’s garbage.  If you want to charge for something we a customer takes it, you need to pay them if and when you take it back.

Don’t Make Your Customer Loyalty Program The Best Reason NOT To Use Your Service

As customer experience consultant Shaun Smith likes to say, “customer loyalty programs don’t generate customer loyalty.”  What they do generate is a lot of system-gaming and one-dimensional customer evaluation.

I remember shortly after I moved to Paris many years ago, I had a full-fare business class ticket from Paris to Tokyo on Air France.  When I tried to check-in they said they were going to downgrade me to economy because I was the least frequent flyer in business class.

I call this the “dark side” of CRM.  Companies start segmenting customers by some variable that they hope correlates to a key business driver like profit or revenue.  This then has two effects.  Regardless of their go-forward business value:

  • People who historically score poorly on that variable are treated poorly, creating a huge disincentive to join the system.
  • People who score well on that variable are treated well and quickly develop a sense of entitlement, creating inflated expectations that are difficult to meet.

Frankly, I liked Virgin America a lot more before they had their frequent flyer program because they used to treat everyone well.  Now, they’re discriminating like everybody else.

Admit Fault When Due – And Act Accordingly

As we endured our nightmare, after many argumentative exchanges, the agents with whom we spoke eventually determined that the person who put us back on the earlier flight had made an error in somehow not properly re-issuing our tickets.   So, while they were not quick admit error by any means, they eventually did.

However, what happened next was appalling.

“Yes, I can see the problem.  One of our agents made a mistake.  And there is nothing I can do to fix it.”

The prior sentence is one that should never be said in your call center.  If you know the mistake is on your side, then you must do something to fix it.  Period.  Always.

Be Mindful Of My Costs In Dealing With You

Because we called the US support number, we were paying in excess of $1/minute in order to speak with the call center agents.  The US agents pay no heed to this — despite well knowing we were in France.  Worse yet, they’d periodically pop back in saying “bear with me” and turn off before we could say anything, so we couldn’t even make an educated ROI decision on whether we wanted to keep holding.  I’m guessing of the 5 combined hours we spent connected to call centers that 70% of it was on hold.  (That’s $210 on hold out of total costs of $300.)

Only when dealing with the French call center — where it customary to pay high telecom fees to be connected — did the agent show sensitivity to our costs, offering to call us back instead of putting on hold.  Thank you for that.  Companies/agents should do it more.

Empower Your Agents To Fix Problems, Particularly Supervisors

It’s well known that both agents and customers are happier when agents are empowered to fix problems.  By my math, our problem could have been fixed for about 300 Euros ($420), assuming they’d have to bump/compensate two people to the later flight in order to honor our committed reservations.  That could have been done on a volunteer basis so the recipient would be definitionally happy.  We would have been happy.  Everyone would have been happy.

Instead, we spent $300 on phone connection fees.  Air France spent 5 hours worth of agent time, call that $300 more.  Air France had to then buy 4 tickets on another airline, call that $800.  In addition, the Air France social media team had to deal with me and my tweets (and delete my post off their Facebook wall).  All in, Air France spent $1100, we spent $300 and nobody was happy.  And they might well have lost a customer for life.

Perhaps you can’t empower every agent to re-book people onto their original but now sold-out flights.  But you can empower your supervisors.  Otherwise, all your agents can do is say no and explain why they can’t fix problems.

Have A Social Media Support Team with Supervisor++ Powers

“If you make customers unhappy in the physical world, they might each tell six friends. If you make customers unhappy on the Internet, they can each tell 6,000.” — Jeff Bezos via the blog

One of the important parts of the social enterprise message at is reflected in the Jeff Bezos quote above.  It’s a key reasons why people also say things like “service is the new marketing.”  The fact is, in a social media world, when people are unhappy they can tell lots of other people very quickly — and that’s precisely what I did, both as a way to vent as well as to test if Air France had its social media act together.

While I wouldn’t give Air France an A in social customer service, I would give them a solid B+.  While a bit slow on the first round, they did reach out to me via Twitter direct message, and even tried to call (but after I was “done” for the day on trying to fix the issue).  So while they were unable to help on the segment 3 issue (and I’m ergo unable to determine just how empowered they are), they did manage to fix the less serious problems I had checking in on segment 4.  I received Twitter direct messages about my problem, they fixed it (overnight), and the even called me the next morning (think:  multi-channel service) to make sure I knew they had done so.

af twitter

My friends at the Service Cloud at figured this out long ago.  Customer service organizations need to integrate social media into the contact center, they need to use classical CRM (e.g., loyalty tier, dollars) and social media metrics (e.g., influence, Klout) in order to determine case escalation rules, and then get high-impact and/or high-pain customers to top-empowered agents so they can resolve problems quickly and turn unhappy customers into loyal fans.

To be clear, Air France did not do that to me.  The modest help with a modest problem did little to offset the frustrations and mishandling of our segment 3 problems.  But it was nevertheless encouraging to see that, in social media support at least, they were taking a few steps in the right direction.



[1] Some of the best customer experiences I have had in my life were in France and largely uncorrelated to price/cost.  Yes, the Tour d’Argent has stunning service, but I’ve also had tremendous experiences at tiny country bistros or cafes.

[2] The net result of which was (a) a ruined afternoon the last day of our vacation, (b) several hours on the phone with Air France / Delta, (c) paying about $1/min for those several hours, (d) numerous attempts at blaming us for what was clearly an airline error, (e) a large amount of frustration including the obligatory argument with my spouse, (f) a failure to honor our reservations for 4 of the 6 people in our group on segment 3 of our flight, routing them on another airline to another airport, (h) doing the prior point despite ultimately having empty seats on the aircraft, (i) uncompensated downgrades from premium economy to regular economy, and (j) several more hours wasted when problems repeated themselves on the check-in for segment 4, the final return segment.

[3] Air France never actually notified us of this change.  We noticed it only after a new person joined our group, bought her ticket on the same flights as we had, and then we discovered the discrepancy.  We then called Air France, were assured that everyone had been transferred back to the 2 PM flight, now flight 97, received multiple email confirmations that this was the case, only to discover on trying to check-in that flight 97 was “sold out” and that half of us had no reservation whatsoever and half were on flight 99 at 8 PM.

[4] While I think this is now discontinued, back in the 1990s they had a really weird configuration where seats on the right side of the plane could slide their arms towards the middle creating a 2-3 configuration on what’s normally a 3-3 plane.  Ergo, the lucky folks on the right actually got wider seats for their money whereas those on the left got nothing.

[5] Lest the relevance of this not be obvious, we were paying premium economy in conjunction with our business class long-haul segments and, due to my size, I really appreciate, need, am willing to pay for, and indeed paid for a larger seat with more legroom and/or the ability to get a bulkhead or exit row.  Having your reservation nuked and squeezed back in at the last minute is the recipe to get seat 29B which is where, indeed, they put me.

[6] I believe that United has been moving in this direction for the past several years.