Testing Tracking
To start off with, tracking does break. It happens. It’s happened on programs that I’ve run and a long time ago (in a galaxy far, far away) I even managed to break some buy.at tracking as a developer.
So the question that we have to ask ourselves is if enough is done to check tracking from the network perspective.
Sure - a good affiliate manager should be able to notice a problem quickly enough, but not all programs have good affiliate managers.
The networks make a lot of money and sometimes have handfuls of account managers, but how much effort would it be to have a small team of people who’s role in life is to place orders online and make sure that the tracking is working across all their programs?
Lets look at some figures to use as an example. If a network had, for example, 400 Merchants on board.
A tester should be able to place around 4 orders an hour - 8 hours a day - 5 days a week. This means that one tester would be able to test the tracking of 160 programs every week.
Get a couple of testers in and networks would be able to check every single merchant program once a week.
Whilst this may not eliminate the problem it may well mean that errors are tracked slightly faster and tracking is fixed quicker - resulting it more money for both the affiliate and the network.
So why isn’t it done? My personal opinion is that networks tend to rely on Affiliates to warn them of problems with tracking - yet again meaning affiliates are becoming consultants working for free to help the networks gain the mega bucks.
I’d certainly welcome a network that had a more proactive testing department and I’m sure that something like this would encourage affiliates to switch their links across.
March 3rd, 2007 at 7:45 pm
There are a few automated services that could do this. One I’ve used is siteconfidence.com.
Basically it follows a set of steps on a schedule and looks for certain results each page. Its normally used to check that your site isn’t broken.
It’s not cheap, but should be well within the budget of a major network and it’s much cheaper than the manual method.
March 17th, 2007 at 12:18 am
What you’re talking about is eminently doable James.
Here at SciVisum we run 24/7 website monitoring/ load testing of complex multi-step user Journeys for clients from the likes of Tesco to William Hill down to small sites at Govt departments or etc - and we run journeys to help clients see the effect of portal performance on their landing page journeys to sales conversion rates.
But the hard things in terms of specifically putting orders through networks in the way you’re thinking are things like: effort in unravelling the stats of the fake orders; for acuracy 1 sample per merchant per week is not enough, so would be placing many more orders than 4 per hour.
Automated payments means reuse of the same credit cards - and hassle from credit card companies; who pays the CC!
But hey, none are unsurmountable - mostly it just requires willingness from the network… I’ve talked to a few in the past, and didn’t get much traction; but maybe it’s an idea whose time is coming.
Be interested to take it further if anyone else thinks so.
Deri