The interesting thing about data quality is it's a mess.
It was a mess in the analog times before,
but you couldn't see it. And now
in the digital world we have
much more data and now
we can see how messy the data
is. Right. And when you look at census data
before on paper or CRM databases,
right now, it's a lot of old
data in there. And with analytics,
data, behavioral data,
we see how messy the data
is and that won't go away.
And especially if we're
talking omnichannel, if we're talking
marketing, that's data from different
sources captured
in different ways on different
websites. So that's
tricky and we can't solve it.
We can only manage it. And that's
interesting part
because, I mean, it's a three
sided answer, I would say.
First, I mean, we can, of course,
make sure that the that the tracking
that the collection is okay.
We can test we can use tools like Observepoint, etcetera
to make sure the data quality there is great. But second
thing is we always have
to check the data for anomalies
because we will track data
that's not okay and then we need to spot
it. So pattern recognition.
And the third thing is we have
to live with the uncertainty. We need
to know that the data is not okay
and we need to know in which ways
it's not okay,
and then we can try to correct
for that in our analytics part.
