Archive for October, 2014

Right And Wrong Ways Of Using Case Studies

Friday, October 31st, 2014

All business development teams understand the importance of case studies in B2B technology sales. But, when it comes to producing and using them, we’ve come across two diametrically opposite approaches followed in many IT companies:

  1. Kool-Aid: Use case studies for everything
  2. Phobia: Shun case studies altogether 

Vendors in the first category launch into case studies at the drop of a hat. Prospect asks, “what can you do for us?”, they hand out a case study. Prospect wants to know, “how is your solution future-proof?”, out comes a case study. So on and so forth. In short, these vendors use case studies as their only marketing collateral. This approach often fails.



TOC for Capability Document

Because, by definition, a case study is a description of “what we’ve done”. That is, it looks at the past. A prospect looking for a solution today wants to know how their technology partner can support their needs tomorrow. They won’t be satisfied with what the vendor did yesterday. Therefore, even the most well-written case study of the best project executed in the past won’t convey what the vendor can do in future. That’s why a case study won’t cut it at this stage. What the vendor needs is future-facing content like product roadmap, capability documents and thought leadership articles. You can find samples of such marketing collateral on our Pinterest Pinboard.

Once they’ve established their capability via forward-looking marketing collateral, vendors can use case studies to support their credentials. Their case studies are unlikely to mirror their exact offering. That’s okay – their competitors won’t fare better. No one can have five years experience in a one year old technology, if you get my drift.

Moving on to the second category, some companies hold themselves back from developing case studies for any number of reasons ranging from NDA and solution is not yet live through to customer is non-referenceable.

Almost always, we’ve been able to work around these restrictions and help these B2B tech vendors overcome their case study phobia. Specifics vary from case to case but our approach is grounded on the universal truism that’s so well expressed by Michael Lewis in his latest bestseller, Flash Boys: “Speaking about someone in a positive light does not violate the terms of not speaking about someone at all”.

In short, case studies can be effective but they need to be used judiciously with other types of marketing collateral to deliver maximum bang for the content marketing buck.

What The Obama Credit Card Decline Means For The Future Of Analytics

Friday, October 24th, 2014

President Barack Obama’s credit card got declined in a New York restaurant. Because, to quote Obama himself, “…I guess I don’t use it enough, so they thought there was some fraud going on.”

While this is the most high profile example of predictive analytics going awry, it’s not the first and it certainly won’t be the last. Many of us have gone through the experience of our credit cards getting declined, which happens most often while we’re traveling, making an outsized purchase, using the card too frequently – or too infrequently, as in the case of President Obama.

dmpa03When a genuine cardholder’s transaction is declined, a “false positive” has occurred.

In a more generic scenario, the term refers to the condition in which the Predictive Analytics software raises an alarm which turns out to be false. False positives have been a major bugbear of Predictive Analytics technology ever since their inception. And, almost since then, we’ve been hearing assurances from practitioners that False Positives would decline as predictive analytics systems gathered more data, made more predictions and became more “self-learning”.

If someone believes that claim after the not-so-new technology just bungled with “the most powerful man on the planet”, well, I’ve the proverbial beachfront property to sell them in Arizona, London or New Delhi.

False positive is a major reason why, in my last blog post titled Difference Between Data Mining And Predictive Analytics, I’d rooted for not spending your entire analytics budget on Predictive Analytics and instead allocating a part of it to Data Mining.

Agreed that Data Mining appears a bit unscientific. But I suspect that’s more the result of following the purist approach of mining all kinds of data and letting the “chips fall where they fall”. The problem with that approach is that chips often fall wrongly or, even when they fall correctly, they fall on the wrong heads or at the wrong time or both.

Data Mining can get rid of its perception problem by focusing on areas that intrinsically make business sense.

Let me illustrate this with the following real-life case study of the use of Data Mining (locales and other details changed to protect customer confidentiality):

“There’s a significantly higher attach rate of business loans with home loans in Zip Code 23508”

dmpa02This finding made business sense in hindsight when the bank in question discovered that its business banking and retail banking sales people sat at the same office in Norfolk, a practice that led to better exchange of market information.

As we can see, Data Mining delivered a valid business insight.

The team of Data Mining experts followed up with the following recommendation:

“Let’s validate this finding in Norfolk. Then, let’s apply this practice to all our regions. If it works equally well everywhere else, our revenues will increase by 8% with no significant increase in cost or risk.”

It’s now clear that Data Mining also delivered a valuable business insight.

Like in this case, it’s possible to take a business-oriented approach towards mining data. When done right, Data Mining can pull analytics out of the hair-splitting mode that it frequently finds itself in and elevate the function to a strategic level.

While there are many implementation challenges around process, HR and so on, adopting Data Mining is a great way for analytics to gain a seat in the boardroom.

Five Ways To Win Back Lost SaaS Customers

Friday, October 17th, 2014

Come Back to Otixo

Many of us have signed up for free software trials and failed to upgrade our subscription to a paid version when the trial period ended. Some of us might have bought a subscription for the paid version, used it for a few months, and then bailed out.

In all these cases, we become the SaaS vendor’s “lost customers”.

According to common marketing wisdom, it’s 10X more costly to acquire new customers than sell to existing ones. While selling to lost customers may not as easy, this segment is still a low hanging fruit for most vendors.

Ergo, it’s worth exploring ways to winning them back. This is what I’ll do in this post. (Avoiding the loss of an existing customer in the first place is another topic that we’ve covered in an earlier post titled Best Practices For Reducing SaaS Churn).

Some vendors have already started taking rudimentary steps in this direction.

Otixo is a good example.

The company sent me an email recently, announcing a freemium plan, outlining new features and entreating me to “Come back to Otixo”.

When I received this message, I couldn’t place the company. When I finally remembered what it did, I couldn’t remember when or why I’d signed up for Otixo. Not surprisingly, I ignored this email. Otixo failed to convert a lost customer.

This experience made me wonder what Otixo or any other SaaS vendor could do in order to achieve greater success in reviving their lost customers.

Here are five things I could readily think of:

  1. Reintroduce itself with an elevator pitch. Not having used Otixo for over a year, I completely forgot that the service speeds up the process of copying files from one cloud location to another by obviating the need to first download the files from the online source to the local PC and then upload them from the PC to the online destination.
  2. Tell the customer when they signed up for the service. The vendor should be able to find this data easily from its registration database. In my case, it was sometime in 2013. A date or at least the month of initial signup would help provide the customer with the background.
  3. Give a glimpse of the purchase journey viz. (a) How did the customer find the software? (b) Why did they sign up? (c) What path did they take to make the purchase? I don’t remember any of these things regarding Otixo. SkyGlue, Convertro and other visitor tracking technologies can help answer these questions.
  4. Jog the user’s memory with the first action they took upon becoming a customer. Logs could provide this information. I think I’d copied a few files from a website to an FTP zone but I’m not sure.
  5. Finally, entice the user back with an attractive offer. While Otixo’s email did offer a free plan, it didn’t float my boat since I lacked the background – essentially the information mentioned in point # 1 through 4 above.

While there could be many other ways to bring lost customers back on board, the above list should serve as a good starting point.

Happy Reviving!

Difference Between Data Mining & Predictive Analytics

Friday, October 10th, 2014

I’ve been asked several times about the difference between Data Mining and Predictive Analytics. Well, that’s not strictly true. I’ve been asked only once and that too in a tweet from @IBMAnalytics addressed to me among 15.7K of its followers.

Replies ranged from the ubiquitous “depends” – do some people always begin their answers with this word? – to “Predictive Analytics is a form of Data Mining that does blah blah blah”.

Since I wasn’t too satisfied with any of these definitions, I came up with one of my own:

Data Mining uncovers hitherto unknown relationships between measurable variables whereas Predictive Analytics predicts qualitative outcomes from measurable variables.

There’s a lot of mathematical theory behind these two branches of analytics / business intelligence dating back to several decades. But, since this is a marketing blog and not a mathematics journal, I’ll skip all that and illustrate the difference between the two technologies by using a few real world applications.

Data Mining (DM)

Sales of cigarettes and detergent rose and fell in tandem in Norfolk, Virginia. 

  • dmpa01Sales of both products are measurable variables.
  • No one asked the DM software to find any relationship between the sales of these two products. In fact, apart from the most clairvoyant amongst us, no one would even suspect that there was any such relationship. I can hear marketers and brand managers say, “Cigarette and mint sales could be related. Ditto washing machine and detergent. But how can there be any relationship between cigarettes and detergent?”
  • The DM software crunches sales data of many products – not just cigarettes and detergents – in many supermarkets in Norfolk to uncover this hitherto unknown relationship between just these two product categories.
  • Even after they receive it, the software’s insight might not make much business sense to marketers and brand managers, let alone their bosses. At this stage, they’d be asking themselves, “Why’re sales of cigarette and detergent linked?” This is typically how DM operates. It focuses on correlation (“what”) and not cause-effect (“why”).
  • DM operates on aggregate data gathered over a period of time and not at the level of an individual transaction or in real time.

Predictive Analytics (PA)

John Doe’s credit card is now being used fraudulently at D-MART Mumbai.

  • dmpa04Fraud is a qualitative outcome.
  • The PA software starts with geographical location, spend velocity, category spends and other measurable variables related to John Doe, the cardholder in question. It then computes an index, based on which it qualitatively flags off the transaction as genuine or fraudulent.
  • Predictions are generally the result of cause-and-effect relationships.
  • The PA software starts with the following inputs (“cause”): (a) the cardholder’s billing address is Norfolk, Virginia, USA (b) the last genuine transaction on this card happened in McCarthy Mall in Norfolk two hours ago (c) It takes at least 14 hours to travel from Norfolk, USA to Mumbai, India. By joining these dots, the software infers that John Doe couldn’t have reached Mumbai from Norfolk so soon. This leads to the inference (“effect”) that John Doe’s card is being used by some fraudster in Mumbai.  
  • PA generally operates on individual transactions and in real time.

Based on the above illustration, if you’re inclined to believe that Predictive Analytics is “more scientific” than Data Mining, you’re not alone. Despite attracting several highly vocal flagbearers – like Tibco CEO Vivek Ranadivé – during its long period of existence, you’d be hardpressed to find too many applications of Data Mining in the business world.

However, that shouldn’t mean that you should forget about Data Mining and divert your entire analytics budget to Predictive Analytics.

I’ll explain why in another blog post in future. Spoiler Alert: False positives in Predictive Analytics; the “two second advantage” delivered by Data Mining; making analytics strategic.

How To Fulfill Targeted Offers To Hear More Ka-Chings!

Friday, October 3rd, 2014

All of us receive spammy offers from banks, consumer brands, ecommerce websites, retailers and many other businesses. If you’re like me, you probably delete these SMS / email messages without even opening them.

Once in a while, we get a discount coupon on our birthday or wedding anniversary. They are more targeted but, since human beings are conditioned to receiving gifts – not discounts – on their Major Life Events, they meet the same fate as spam.

It’s rarely that we get an offer that makes total sense to us when we receive it e.g. 20% off on Coke just after buying a pizza; offer to restructure an atypically high value credit card purchase into six equated monthly installments, and so on.

The recent launch of Apple iBeacon has raised the potential of – and the buzz around – targeted offers to the next level and we should be hearing more about them going forward.

Now, cue to what happens when we decide to partake of a targeted offer.

You go around in circles. Like I did on at least three occasions in the recent past:

  1. A taxi hailing app sent me a discount code that I wouldn’t be able to remember whenever I ordered its cab next. More on this at Walking The Tightrope Between Driving Repeat Purchase And Rewarding Loyalty
  2. One offer app installed on my smartphone correctly recognized my location (Kalyaninagar, Pune, India) and showed me a 30% discount offer for a burger from a nearby restaurant, with delivery charges waived for good measure. When I tapped on the REDEEM OFFER button, the app took me to the quick service restaurant’s website to place my order. However, the QSR thought I was located 2500 kms away (in Gurgaon, India, BTW) and politely declined home delivery since my address was outside its 3 kms delivery radius. To add salt to the wound, the QSR informed me that the offer had expired five months ago.
  3. Another offer app sent me two targeted offers in a certain mall. When I tapped the offer’s PUSH notification on my Android’s notification bar, the app generated *ALL* offers – around 50 – in the said mall and lost me in the resultant clutter.

Going by these examples, many companies seem to be paying scant regard to the crucial process of offer fulfillment, which is a pity considering the time and money they spend on making them.

lto02But not Go Daddy.

The world’s #1 domain registrar recently sent me an email with an offer for 20% discount on my next order. Since I was in the market for a new domain name, the offer caught my eye immediately. My interest went up a notch when I noticed that the promo code given in the email was clickable (yes, I notice such things). Therefore, I went ahead and clicked the link. I was taken to Go Daddy’s website, where I logged in with my credentials.

Now, if you’ve shopped for domain names, you’d know that the process comprises of several steps viz. (a) enter the domain name (e.g. (b) check availability (c) select the term (e.g. 5 years) (d) enter nameserver details (e.g. NS183.HOSTGATOR.COM) (e) punch in the credit card details (f) add promo code, and, finally (g) click the BUY button.

Go Daddy didn’t ask for my promo code at any stage. Instead, the website applied it automatically to my checkout and assured me that it had given me the best price. I whipped out my calculator (yes, I do that a lot, too) and satisfied myself that I’d received the promised 20% discount.

Behind the scenes, Go Daddy had remembered – or “persisted” in geekspeak – my promo code.

Now, persistence is not a big deal. Many websites do it. For example, I’d highlighted in If You Must Use A Long Form, At Least Pre-Fill As Much Of It As You Can how a click of HubSpot’s Promoted Tweet for a webinar took me to a microsite where the registration form was completedly prefilled. However, barring Go Daddy, in all examples I’ve come across of persistence, the data is persisted only on the landing page immediately following the click.

However, Go Daddy persisted the code well past the landing page, which is unique in my experience.

Kudos to Go Daddy for going the extra mile to implement “multistage persistence”, which delivered one of the best fulfillment experiences of a targeted offer that I’ve come across. I’m sure the company is reaping a rich harvest by enjoying above-average redemption rates for its offers.