My problem with data analytics

As a career marketer, I find analytics endlessly fascinating, so please don't get me wrong here, but the number of B2B companies investing in analytics worries me. Not because of the analysts themselves, because I have met and worked with some very talented individuals, but because of the data they have to work from.

It is not perfect, by any means, as regular readers of this blog will have heard me explain a good few times. Unlike consumer analytics, where assumptions can be made about people who live in certain postcodes etc, businesses do not fit any particular pattern, and we don't know enough about them to really segment with any confidence.

Pick any B2B data universe, and I could pick holes in it. From the very top all the way down to the murky depths. Compiled datasets with legal registration data matched with directory data, maybe a little tele-research and if you are very lucky a bit of transactional information does not give a very solid base to base any analytical project on. But these datasets are rightly considered fit for purpose (in most cases) and they represent the best we have to work with.

So unless a large brand has a very good CRM database to add extra insight to licensed data, any analysis will be based on a number of assumptions, matching processes and Companies House information. Read some of the analytical boys' websites and you would think they can find a needle in a haystack, or even a laptop on a commuter train, but I am not so sure.

I also have my doubts about Santa and the tooth fairy, by the way.

We are all doomed, doomed I say

Luckily, I can't type in a Scottish accent, but for you young one's out there, the title of this piece is a reference to a famous catchphrase in Dad's Army, always said by Private Frazer as the episode went *** up. The reason for my pessimism today is consumer confidence in direct marketing and selling. It's dead and gone basically.

I've been assisting a data collection project in the last few days, and I did some door knocking myself. Out of several hundred addresses visited, 75% had signs up banning direct selling and 30% 'junk' mail. Quite a few had signs banning free newspapers. Sometimes you couldn't see the door for the stickers.

Door to door is no longer popular. I am old enough to remember the man from the Pru calling each month, and milkmen, and countless other tradesmen who called door to door. Now they would not be welcome. No one wants the intrusion, no one has the time to waste.

So what, times have changed, I hear you say, and you have a point. So much of life can be done online, who needs door to door services, who needs Post Offices, who needs post, or any human contact at all.

Yesterday I mused on consumer confidence in giving out their data being damaged by all the theft and loss stories we've had recently. The other side of the coin is that the really direct approach, knocking on doors is out too. Door to door DM offends the green's and the consumer, because they hate having to recycle unwanted junk.

Something has got to give. We have to find a way of communicating with people in a safe, green, friendly, interactive way that does not intrude or offend, otherwise direct marketing (in it's broadest sense) is doomed. As the recession starts to bite, a lot of businesses are going to start to struggle. There will be blood on the walls.

And they still don't like it up 'em you know, Captain Mainwaring.

Am I missing something here?

I am not a regular commuter, thank God. I find being crammed into an overcrowded carriage and usually denied a seat despite paying a small King's ransom for the honour quite distasteful. However, about twice a month I do make the trip and I am yet to find all these laptops being left unattended.

It seems to me that every other day some poor civil servant is leaving his laptop on the train. Either that or losing it in some other negligent or criminal way. I had a pleasant chat with Gemma from Marketing the other day, after that survey made it onto Radio 4. 61% of marketers have admitted to losing data, or having it stolen. My assumption was that most would be the former.

We have to wise up, all of us. Most organisations, including data owners, suppliers and brokers, treat their data with a disdain bordering on the disinterested. Consumer confidence is taking a hit every time one of these stories hits the headlines. If we cannot turn it around and start treating customer data, and the DPA, with respect, we are going to find it increasingly hard to source fresh data, or hang onto what we already have.

The database is an asset, so we have to invest in it, and protect it, like any other asset.

Hope old Gordon checked TPS

I once wrote to Tony Blair, about taxation, as I remember. It got passed down the line, and some lowly bod in the treasury replied with some guff about it all being for our own good. Thank goodness Tone didn't ring me up. I might not have been able to contain myself. So the thought of Golden Brown giving the odd voter a ring to 'discuss' things with them is probably worse.

According to Fleet Street, he even called quite early in the morning. Not good Golden. No point in deliberately cheesing off the punters with an unsociable call. You will never get return on investment that way. And I hope you got your office to run the number(s) against TPS. Wouldn't it be embarrassing if you made an unsolicited call?

Of course, you could argue that as the voters in question has written to you, you had a right to call them. Well, no, it doesn't work like that. They probably wrote complaining, as I imagine very few people ever write to the PM praising him (in general, I mean - in your case, probably not at all!). They wanted you to actually do something, not wake them up with the dawn chorus, not talk about long-term decisions blah, blah, blah. Actually do something.

To suppress or not to suppress

More debate in the weeklies this week about the failure of so many mailers to use suppression files. I am afraid the hand-wringers are missing the point.

Fact one, suppression costs money, and time, which costs more money. Yes, it saves the cost of wasting postage, but if new owners have moved in, you never know, do you?

Fact two, there is little downside to not using suppression. We are all so ennured to receiving junk mail we barely shrug as we toss it in the bin. Brand damage potential, maybe, but nothing hits the bottom line in the short term.

Add one and two together, and you get general apathy. A waste factor once the campaign has mailed will be irrelevant if the campaign achieves the desired ROI. So no one cares.

Yes, this is wrong. Yes, we should do better, but do we trust the suppression files themselves?

I am afraid it is all about the bottom line. The costs might seem marginal to a large business, but the truth is the budgets may not match the size of the business. Everyone is penny pinching and if there is no penalty for not using suppression, the cost argument will win.

As for the suggestion that the DMA might start throwing out members for not following best practise, I am still laughing about that one...

Being positive about B2B

I always am, really. Positive, that is. But an article in PM last week did catch my attention, as the suppliers spin doctors got to work. So, an attempt to see through the smoke and mirrors from B2B101.

Undoubtedly there are suppliers trying to be more sophisticated with B2B data, but the article was a little lop-sided. Only Vic Godding put the other side of the coin.

Vic pointed out that the true cost of improving quality was a deterrent to actually delivering better quality. The facts are that the base data every database is built on has inherent problems. The various bits of information from Companies House, directories and subscription data from various sources has to be matched to give a true mailable record.

So in that first process there is a built-in error percentage. What is it? Your guess is as good as mine, I suspect, but I would say 10%. Then you have to keep those records up to date.

Well Companies House is basically updated once a year. Directory data is updated less than that, around every 18 months on average. And subscription data is also annual. In 12 months, industry folklore suggests a 40% decay rate. So you can argue that you can question half of all data on the market right now.

Tele-verification is expensive, and that is why only small portions of the universe are ever contacted by telephone. Returns and suppressions are applied in the best cases, but it is only ever papering over the cracks, as decay works faster than updating.

So, analyse that, chaps. Yes, you can introduce sophistication, and develop highly intelligent products, especially if you focus on a small subset of the universe, but anyone who argues that an entire universe can be treated in the same way is barking at the moon.

What marketers need, and expect, is fit for purpose data. A supplier should be able to offer a selection or a means of segmenting that approaches client expectations, but we also do ourselves a disservice if we cloud the sales process in smoke and mirrors.

The PM article contains some implied criticism of job titles. But the truth is that is all we can be certain of. Who is actually holding the position at the time may sometimes be beyond us, since people do not move on at twelve month intervals, but we can be sure there will be an MD, an FD, a sales manager and a marketing manager.

As much as I dream of the day when we can be truly sophisticated and target much more effectively, the reality is we cannot. The ingredients just aren't there. If a supplier is using more smoke and mirrors than Britain's Got Talent, try asking them some key questions. How much do they spend on verification? What is the updating cycle for their data? How do they collect it, what suppressions are applied, and when, and how do they handle returns?

If the answers sound like dear old Golden Brown (We are making the right long-term decisions, we are getting on with the job etc etc), rather than straight, honest, factual responses, I would think twice before placing an order, if I were you.

I was trying to be positive, and there is good data on the market. It is a horses for courses situation, and you will pay for better quality. It is not all doom and gloom, but it is not all sophistication and insight either.