Today, 25 September, is my birthday and this reminds me of something:
Data Collection Efforts That Invite Crap Data
Crap data leads to questionable analysis, bad business decisions, and even contempt of your own customers because they’re the ones who fed you this hogwash.
It’s often said that our lists are the most valuable element in our businesses. Contact lists, client lists, prospect lists, webinar participant lists, etc. These lists are often the only thing that’s exchanged when someone sells a business. How about the data quality, though?
One data collection effort that is a magnet for poor data quality is online forms with Required Fields, specifically when the Required Field is personal and is not germaine to current purposes. Requesting and requiring information in this context is technically called OFDLIZ:
open front door,
let in zombies
Sell a business and exchange crap data for money? :facepalm:
BIRTHDAY as a required field is something marketers need to think seriously about. Birthdays are undoubtedly valuable information TO MARKETERS. But marketers need to give us a damned good reason to provide the truth, and they have to gain the trust of a public that’s wary of identity theft.
I am regularly guilty of inputting an incorrect birthdate when the marketer
- doesn’t need my birthday
- I just want access to read an article about Chicago’s plan to make city parks wi-fi.
- only needs to know that I’m over a certain age
- To use social media sites
- To access a site and read reviews on bourbons and ryes
Here’s a question someone may ask: Well, Mr. DataMan, how do you know that the birthdays are crap?
ANSWER
I was helping someone scrub 1000+ rows of data submitted by people who’d registered themselves online. A surprising number of people were born on 1/1/11, 6/6/66 and 1/2/34. Interesting. And plenty of “555” phone numbers.
This leads to other questions:
- How many people abandoned the form/registration because it became a hassle?
- Were ANY of the 6/6/66 birth dates real? That looks more plausible than 1/1/11
- How many of the normal-looking dates are crap data?
- Why wasn’t there sufficient validation to prevent 0/0/00 as an acceptable date?
- Is the birthdate easy to revise if a person wants to provide the real date later?
SUGGESTIONS
- Marketers should gain our trust. State how this information will and will not be used.
- Only require fields that are necessary.
- Required fields can make a form a hassle to fill out. Pressing ENTER and having the form come back with red boxes is un-fun.
- If you want to send customers a birthday email, then just ask for the date and month. Leave the year optional.
- You can ask for anything under the sun (eye color, medical history, dietary restrictions, political beliefs) but leave the option open to users to provide it if it’s only for marketing purposes and is tangential to the purpose of the interaction.
- Explicitly invite users to restrain themselves from loading crap data. Invite their empty fields. Love their empty fields.
Of course, all of this means you’ll get less data from your audience but you’ll be proactive in increasing the data quality. Otherwise, sit down and wait for the zombies.
Additional Reading: Using New Data for Marketing That Is “JUST RIGHT” Eric Wittlake raises moral and ethical questions about advancements in technology and how data can be used in marketing to us. This triggered my concern about data quality, something that Eric is also concerned about.
doorway photo credit: maistora via photopin cc
lightbulb photo credit: Raymond Larose via photopin cc
zombie silhouettes photo credit: Ben Blogged via FreeVector
hahahaha! You definitely got my attention with those zombies! Yes, I’ve never quite understood why marketers create high barriers/hoops to jump through, especially when they are attempting to attract new leads. The more information I have to give someone with whom I don’t even have a relationship (and when it’s irrelevant to the offer), the less likely I am to give them anything……or, like you illustrate so well, they will end up getting incorrect, worthless info!
tshombe beautiful hearing from you. Thanks for the comment. Glad to know you can identify. And I appreciate your commitment to successful businesses and entrepreneurship through your radio show http://www.spreaker.com/show/selling_with_spirit_radio Have you done shows on the need to maintain trustworthy data?
OzData tshombe You know, Oz, I haven’t done anything on the need to maintain trustworthy data. BUT, I am most definitely intrigued. Am I being to bold to suggest that you might like to be a guest delivering expertise on this most-important topic?My audience is highly concerned about the know-like-trust factor as it relates to client attraction and retention. Having inaccurate data on prospects, leads and clients and in turn making business and marketing decisions based on that data engenders neither “like” nor “trust.”Sounds like an awesome topic to me!
WOW! I like how you’re thinking. (This is humbling.)I’ve had clients whose businesses were nearly paralyzed because their data wasn’t clean (filled with duplicates, incomplete data, inconsistent formatting, info spread out between various spreadsheets and CRMs). After a while, a business is running on educated guesses. And it’s not because people are incapable or lazy, I think it’s because Data Management isn’t a clearly-defined activity that has to be tended to.And yes, “like” and “trust” go out the window when a person is “known” as “that company that sends me 3 of the same thing because they can’t purge duplicates out of their database.”It would be great to share some expertise in this area. I’m flattered that you asked. So, let’s do what we can to keep businesses running smooth, sustainable and profitable. tshombe OzData
OzData You know, Oz, I absolutely love your passion and expertise about helping entrepreneurs in this area. Just today, I received another mailing (direct mail) from someone I follow and respect who has duplicate information on me. The first post card was mailed to Tshombe Brown and the other to Brown Tshombe. This is not the first time this has happened with this same provider.Another example: I am on an email list where somehow they got ahold of my middle name. So now, every email I get from them addresses me by my middle name rather than my first name, even AFTER I specifically pointed out the error to them!Having inaccurate records is the fast-track to losing credibility.I love how you offer that these mistakes and inefficiencies have nothing to do with being lazy or incapable (I noticed that you have said this before on your blog), but rather a lack of attention to data management. The fact is that most of us have no idea that we should be attending more carefully and specifically to data management, its implications when we don’t, and even how to go about it in a systematic, routine way.I’m grateful we’ve connected and look forward to working together to support businesses to, as you so eloquently put it, run “smooth, sustainable and profitable.”
Interesting. Now imagine the entrepreneurs who refuse to do a mailing because they know that what happened to you is likely to be the result? With crap data, the choices are:1. Do nothing or,2. Do the mailing any way, waste postage and frustrate some clients/customers.Depending on how the data is set up and where it is, correcting these multiple mailings can be corrected in a fairly short amount of time.It could be a download into Excel. Write formulae to flag obvious duplicates. Purge the ones that are to be dropped. Set the clean ones aside. (Don’t use Excel’s duplicate finder because it picks the duplicate to purge and there is no un-do. It might purge the record that had your company name, and kept the record that doesn’t).Then things get tricky in order to extract slippery (non-obvious) duplicates (e.g., the variations on your name). Maybe we make a unique identifier out of the first 4 characters of your zip code, and the first 3 characters of your street address. That may provide something unique enough to sort and centralize every record that’s you, and with the different variations on your name.Let’s say you live at 107 Green St. Zip code 38228The unique identifier would be3822107We sort by the unique ID and create flags for duplicates.Some people won’t understand what I just typed but the point is that there are1. Solutions to cleaning such data, and they do not involve eyeballing thousands of records one by one.2. Prevention methods that can minimize a lot of this at the entry pointBut this is like a lot of things: it can do wonders just by making people aware of solutions so that they’re empowered to ask the right questions in the right arenas. But until that point, it just looks like a monster.So, yes. Definitely, let’s arrange to talk Data Management with your audience. tshombe OzData