The last few posts of mine have talked about various tips, techniques, and traps of using your data — be it “Big” or just plain old regular-sized data. This month, I thought I would toss into the conversation a topic that brings smiles or cringes, depending on your take on the subject.
Yep, I will go there. I hope we can navigate the topic and still sit down together and break bread someday.
So, what exactly is a recommendation engine? Think of it as the computerised version of a personal shopper. Like a personal shopper, the engine learns what you like and picks similar, or related, things for you to ......[more]
27 March 2016 · By Greg Bright
Your first big campaign is in the books. Great team effort. Great response rate.
On to the next campaign. Same great responses. And so on with each campaign, getting an OK response.
But, if you look back at campaign No. 1 — which you ran four years ago — to the campaigns you run today, you’ll see a trend. When you look at the long-term trend, you aren’t getting quite the response you used to get. Refreshing the creative reverses the trend … for a little while. You refresh the creative again, and the data people get questioned for their list pulling.
Yet, still there is no way to escape — the slow downward trend is there.
A fear sets in: You are now four or even five years into your data-driven acquisition campaign strategy with slowing response rates, and you wonder if the good old days will ever be ......[more]
01 March 2016 · By Greg Bright
Campaigns that deploy across channels are a fundamental expectation from the executive suite. It is expected that marketing efforts are calculated programmes to deploy a message wherever people are and attract (convert) as many as possible to customers.
And rightly so. But the trick now more than ever is to put the message in the right place at the right time.
Pre-Al Gore (aka, before the Internet) it was straightforward. Now it seems like there is a channel to the consumer born every day. No single channel guarantees success so you hedge your bet and play everywhere — from the traditional trio of e-mail, telemarketing, or direct mail, to the extended channels of outdoor (billboards), radio, or television, or the emerging social or geo-push message and beyond.
For all the talk about disaggregation of media content, advertising is worse! Where do you put your money? Does it still make sense to manage the timing of the message across channels?
Where do you start? Budget? Goals? Audience targeted event anchoring? Message delivery vehicle?
Answer: All of the above.
This isn’t the classic cliché of three elements where you have to pick two assuring failure. You really do need to balance all of the questions to form the answer.
For example, if you want to target apartment renters with a subscription, you need to start with forming an understanding of the audience group. Are you going after all ......[more]
02 February 2016 · By Greg Bright
Last month I talked about the half-life of data – how time not only degrades but can destroy value if not governed carefully. For example, a stop for non-pay followed by a restart then a stop for a move-out-of-area; they all loose relevance when you get a next new next start at the same address.
Assuming the new start is a different name and/or phone number, the value of the previous transaction history for much (if not all) of your standard analysis is destroyed. If not filtered carefully, it can lead you to erroneous conclusions.
Why? A different person is now at the address. You need to track his/her activity separately.
This month, I’ll weave through some of the considerations you need to ......[more]
05 January 2016 · By Greg Bright
There is a rule of thumb in the data world that you should save everything. Every stop, every vacation, every payment. If you can think of it, save it. You might need it. Storage is cheap. Save it all.
Well, sure, you can save it. But does data collected ever reach a point where it is too old to use? Is it possible that old data may be leading you to a flawed analytical conclusion? Is old data setting the stage for errors and waste in your marketing campaigns?
Or, is having ancient information enabling you to discover key insights that would never be possible without the information extracted from the history files?
The answers to the above questions, without a highly refined data governance and usage policy in place, are all: “It depends.” For each situation it is critical to know the context of the use planned for the old data.
Let’s take a quick look at a simple example – a telephone number. The phone number in question is from a customer who last received your publication (or site access) three years ago. Is there value in having this phone number? Should you dial it in a telemarketing campaign?
Well, it depends. You have to look at how the customer associated with the phone number left you. Did the customer leave you with a transaction indicating a move from market? Or did he leave you with ......[more]
01 December 2015 · By Greg Bright
Oh, this post will be fun.
My hunch is that, from the time I write this to the time you read this, one of the vendors will be no more and three new ones will have sprung up. This is the unfortunate reality of the ecosystem of Big Data.
Hadoop, NoSQL, and all of the others living in the cloud storage/tools world are in an explosive stage of development, with many new companies jumping into the fray with a widget that makes something possible (time series) or the large firms making something easier to do (drag and drop H-SQL).
All of this inventive effort makes selecting the right set of tools to build your technology stack very difficult. The classic cliché of 20:20 hindsight will prove yet again that your great tech decision was wrong. But, you can’t stand on the sidelines; you have to pick something.
So, if you are sure you and your company are ready for the Big Data stack (versus lots of data), let’s jump down the rabbit hole ......[more]
18 November 2015 · By Greg Bright
You would think this topic would be pretty simple. At 50,000 feet it is: Split the e-mail list into two halves and call it a day.
In the weeds of implementation, it is a whole different world. How big should the test group be: 50%, 20%, or something else? Should you test in an A/B mode or multi-variate? How much (or little) should you change between the A and B designs?
Should the design look somewhat similar, or is a radical variation ok? What criteria do you use to decide how to split the list into groups? How do you vary the pieces: Is one a true control, or do you just constantly try variations?
In e-mail campaigns, can you use opens/click thrus to dynamically pick the winner mid-campaign? How do you measure direct mail? Are there other campaigns happening at the same time to the same groups (telemarketing, for example)?
So much for simple.
Then when the campaign completes, if the test group sells 50 units, and the control group 43 units, the winner is easy to pick, right? Or is it?
To split or not to split
The short answer: Split every campaign related to sales and retention efforts. Every campaign is a chance to test and improve. People are fickle and will respond differently to ......[more]
28 September 2015 · By Greg Bright
This is a continuation of my previous post, which began exploring the difference between Big Data and lots of data.
Another way to approach the Big Data technology decision is to start with looking at your data and categorising it into what is and what is not Big Data.
It is not Big Data, in my opinion, if you are looking at customer account history, digital subscription access log summary data (device, OS version, user information), or payment history.
Nor is it predictive analytics; predictive analytics is a process applied to data (big or not), not the data in and of itself.
Likewise, if you are sitting on less than a dozen terabytes of data and it is growing at less than 20% a year and you are just running reports to understand what happened with your product, you don’t have Big Data (nor do you have a Big Data technology need).
You have lots of data, but not the kind of data or application need from the analytical discoveries that make the Big Data technologies ......[more]
23 September 2015 · By Greg Bright
It seems like every day I get a letter from a company telling me that I’ve got to get a “Big Data solution” installed or my company will fail. I also get “real world of Hadoop” e-mails – usually links to white papers on why everyone is adopting Hadoop and NoSQL solutions to solve their problems.
The phone calls come as well. Usually the calls are from big name IT firms. They invariably have the same message: They have a solution and consultants that will solve all of my problems by installing a Big Data solution.
I tend to like to play with the callers a bit, first to stay up with the latest technology, and secondly to see if the callers really understand my company or if it is just a cold call script.
I say something like: “What problems? Do you even know my business?” They usually don’t have a clue about my company, making their replies sound like the reading of a script loaded into their ......[more]
27 August 2015 · By Greg Bright
Mistakes happen. The goal with training, checks and balances, and defined procedures is to prevent as many as possible – hopefully all of them.
However, some things inevitably are not thought of, appear out of nowhere, and catch even the best organisation off guard.
To keep mistakes to a minimum, when helping companies launch direct marketing efforts, I give them a list of the top 10 things that will go wrong in hopes that they can avoid everything on the list. I’ve come to realise that many of the mistakes seem to be unavoidable, and each client will make them regardless of any advance warning.
The inside joke is that there are now more than 25 things on my top-10 list.
Between problems with dirty data coming in, mistakes in selections, broken software, or even problems with the type of ink in a printer, mistakes happen. And they result in everything from having a good chuckle (the AMEX logo next to the VISA check box) to a major financial issue (the entire 80,000-piece mailing sent to the newspaper’s office instead of the prospect’s because ......[more]