Tesco and very little helps. Plus Google spam pan cans mail fans.

First of all, sorry for not blogging over the last couple of weeks. I have been maxed-out with work, but as we begin to ease ourselves into the holiday season (large parts of northern Europe are already making their mass getaways) I find I have time to offer a few words of wisdom to the great and powerful.

For a start, let us look at Tesco. For those who are not familiar with the British shopping scene, Tesco is the big boy over here. Even the mighty Walmart, that leviathan of North America, has seen its dollars crash uselessly against the sheer market presence of local goliath Tesco, failing to gain significant market share despite its purchase of ASDA (one of the other major players in the UK grocery business). Legend has it that one in every eight pounds spent in the UK goes through Tesco’s tills. It has been a phenomenal success.

This success has, of course, been built on outstanding customer service. Based on price and convenience and responding rapidly to its customer needs, it is the quintessential case study of how to grow by both following and leading your customers’ preferences.

Like any successful enterprise, having dominated one marketplace, it sought to extend its franchise out of grocery and is now a leading supplier of white goods, consumer electronics, clothes and even petrol, where it has shown the gas companies how to tighten margins but make a profit, a lot of it from cross selling.

‘How the mighty hath fallen’ as the good book puts it. As they push into increasingly unfamiliar territory, Tesco looked at the world of domestic finance as the next frontier, launching an array of financial products, including banking. Originally they were delivering their products on the Royal Bank of Scotland’s platform, but over the weekend of 18th to 19th June they switched to their own. On the scale of things this was a significant if not massive migration - 850,000 customers - but to say everything did not go well was an understatement.

It is difficult to isolate the whole range of problems. Some were down to the normal ‘Pilot error’ of large scale online operations (people forgetting their login details etc.). Some were down to the set-up features of individual browsers (IP9 coming in for a lot of comment). But some were undoubtedly down to data migration issues with some accounts simply not making the trip. There were also corruption problems for accounts that were caught mid-transfer when the online system crashed on the Monday and Tuesday leaving it offline for 12 hours one day and 6 the next.

What happened then was an excellent example of (poor) disaster management. The call centres were overwhelmed with people who could not access their accounts. The store managers were insufficiently well briefed when Tesco cards were being refused at their own check outs or when people came in to complain. In a manner with which I am sure we are all familiar, each fix only served to reveal yet another layer of problems, the consequence of which was that with each day, as the disaster unfolded, promises that everything would be alright in the morning proved so many false dawns.

By the following weekend, the problems were still not fixed and senior executives were appearing for a grilling in the media. A series of interviews followed, not all of which were handled with quite the degree of mea culpa one would have liked.

Classic PR errors were made. The CEO stressing how many people could log on did little to comfort the thousands who couldn’t; in the media, inaccurate accounts of the problems which clearly went beyond the online system to the back office didn’t help. Using a premium rate 0845 number for the help line, then keeping people waiting for 45 minutes, led to accusations that the bank was profiting from the disaster. Pretty much a case study in how to not handle the PR fall out of a data migration failure.

So what are the lessons learned? I’m no PR guy, so I won’t dwell on the right way to handle these kinds of crises, but blaming the customers is probably not one of them. From a technical point of view, first of all it seems that there was no planned fallback (or even fall forward). This was an all-or-nothing strategy with a key constituency of their business. It is not clear what the long-term fallout will be. Tesco stores are just too damn convenient and ubiquitous to hold out against for long, but there is bound to be some short-term negativity.

Secondly, it is not clear how extensively they tested the new system. IE9 is hardly a rare breed of browser, surely they should have seen it coming?

Thirdly, they were caught unaware of the depth of problems their customers were experiencing. This led to appalling expectation management.

Fourthly, it looks like their business engagement structures for communicating with their own staff were not up to the load placed on them.

Finally, it is clear that whatever migration methodology they thought they were using, it failed miserably (it wasn’t PDM). There was not enough in-depth Data Quality work completed. There was not enough soak testing. It seems that there was insufficient band width planned for launch day. The loading was, of course, exacerbated as news spread of problems with the online. What were they planning to do with records that failed to migrate? Wait for people to phone in?

However, having staggered for a few days (well, nearly a fortnight), Tesco, true to its core values, is now responding with a significant air of compassion and compensation. Perhaps they will win their customers back; I expect they will.

If there is any cold comfort to come out of this debacle it is that, for all you folk out there in lonely isolation, locked within the deadly embrace of your corporate rules of omerta, suffering the misery of an equally cocked up migration, you are not alone. Current figures suggest you share your plight with at least 40% of other Data Migration projects.

It doesn’t have to be this way any more.

Those of you about to embark on just such a journey - learn these lessons with care. Public failure and public humiliation are not a good way to build a brand.

Which leads me on to Google. It seems that their Gmail spam filter is tuned too tightly. Here at iergo we work mainly electronically. Orders are received via email, invoices are despatched that way. This last week, Google has thrown a huge spanner in the works by bouncing emails to some of its clients. The client is, of course, not informed. The Gmail web page is littered with acerbic comments from infuriated folks like me who can neither get our mails sent through nor get our names taken off their spam list. Some, bizarrely enough, are Google’s own customers.

So, if you are running your business on the Google platform and have noticed a sudden drop in orders check it out.

Altogether then this has been a bad week for the giants of industrial customer relations. Next week, however, I plan to bring you a story of good Data Migration practice to compensate.

About the author

John Morris has over 20 years experience in IT as a programmer, business analyst, project manager and data architect. He has spent the last 10 years working exclusively on data migration and system integration projects. John is the author of Practical Data Migration.