After my presentation at the eMetrics in Washington, I was approached by an off shore company specialised in data analysis based in India. Or at least, that’s what I thought it was, during our first encounter.
I had already noted the little black note books lying on the tables of the big hall where the keynotes were held and thought to myself that it was good timing as my third Google note book – which fits into my handbag – was already full. But the name of the company did not ring a bell even though the tag lines “data analysis – technology services” spiked my curiosity.
So, when the guy from Theorem actually came up to me, it rang a bell and I wasn’t really surprised by the idea of outsourcing data analysis to an off shore country. India seems the logical step as we had already discussed such possibilities of future evolution within the Web Analytics industry with René when he was first approached by a couple of Indian guys some months ago.
We strongly debated it, at that time, to see whether opportunities of outsourcing web analysis to near shore countries such as Romania for example, as they speak French, or off shore to India or Mexico could make sense in the current context of market maturity we’ve been experiencing over the past 5 years.
To be quite honest, it actually boiled down to defining how repetitive Web Analysis could actually be and whether economies of scale could be found. After all, isn’t that what near & off shoring is all about: doing the same thing for less money, possibly within the same time frame.
I would like to emphasise here the use of the word “analysis” compared to “analytics” as in my eyes, the name of the game of the sector we operate in is finding insights, stemming from analysis.
Web Analytics is therefore a larger concept that encompasses in the end analysis but is mainly currently stuck in issues surrounding setting-up of processes in order to pick-up the right information – often through the tagging method -, display it in the correct format for data consumers to understand and draw conclusions about certain aspects of online communication such as defining whether a campaign has been successful or not, based upon historical data from prior similar activities.
So, how about off shoring web analytics?
The reason why I consider it’s going to be really hard is because in the first place, just getting the processes for tagging for example right is different for every company. From the ones using a centralised content management system for all their online collateral to those dealing with multiple systems and solutions, basic page view tagging is often already a conundrum. And I’m not even talking about tagging for specific requirements such as Web 2.0 type of applications, Flash or external links, amongst many others.
On the other hand, once you get the hang of it, let’s face it, even if most international companies take at least 6 months to get this tagging thing right, it’s not really rocket science on a technological level.
Let’s also not forgot the multiple actors that today still come into play when trying to measure online communication, from media buying agencies to newsletter providers, the agency responsible for the Nth minisite, SEA, you name it.
And all these people need to be convinced you’re not going to steel business away from them, they need to trust you in order to align to the processes you’re trying so hard to make everyone accept and respect, for the good of the client.
Trust me when I tell you that just by sending mails to the different parties involved or talk to them over the phone will not make them fall into line. More often than not, you’ll need to sit into the meetings, go for lunch in order to convince these different parties of the benefits of collaboration.
Proximity holds thus an advantage but some tasks can clearly be preformed from a distance.
Once you’ve got data rolling in, thus you’re picking up the needed information; it has to be moulded into reports, usually rendered by your favourite Web Analytics solution. As we are facing less and less enterprise class solutions out there, getting the hang out of that could indeed become quite repetitive.
After all, that’s also how we started out in the beginning of my little Web Analytics Business Unit: with product specialists, mainly WebTrends, but Web Analytics is today about more than just knowing which check box to tick to render report X or Y.
And in the end, it all boils down to true analysis and giving insights to your clients, which in my humble opinion can’t be repetitive nor benefit from any kind of economies of scale.
I would therefore tend to agree with Davenport & Harris when they note in Competing on Analytics:
“… it is difficult for analysts to develop a trusting relationship with decision makers from several thousand miles away. It is likely that the only successful business models for this type of work will combine onshore and offshore capabilities. Onshore analysts can work closely with decision makers, while offshore specialists can do back-office analytical work. If a particular analytical application can be clearly described by the business owner or sponsor before it is developed, there is a good chance that development of the relevant algorithms could be successfully outsourced or take offshore”.
There is room for offshore or near shore support related to Web Analytics. It will however be quite some time before it is as Theorem Inc.’s short description says “… help[ing] marketers build their business with expert outsourcing technology services and the ability to gain valuable insight into their campaign data”.
It’s the later I’m not convinced about. There’s still something to say for proximity, trust and discussing a client’s strategy while understanding and experiencing the underlying company culture.