For those business stakeholders who can’t make a single decision in their entire lives we have the #1 method for pissing off your entire data team:
Just Ask For Everything
I call them the “Veruca Salt”.
They want the world.
They want the WHOLE world.
The conversation usually goes something like this:
Veruca Salt
Hey guys! I’m working on a big project with a lot of complex moving parts and despite having zero background in working with data, I’d really appreciate it if you sent me all of the data so I can work out what it is I need.
Data Analyst [unfortunate enough to be covering the request process that day]
Hmmm, ok. Can you narrow it down a little bit, anything in particular you want to see? Maybe trends in usage or sales or customer profiling?
Veruca Salt
Nope. I’ve given this zero thought altogether.
Now I'm freaking out because I have an action from a project meeting and really can’t go back with the great big page of nothing that’s on my desk.
So can you just give me everything right now?
When I look at it the real answer will magically appear on my screen like one of those magic eye 3D pictures where a dolphin swims out at you and playfully dances like Flipper in that old 1960s TV show.
Data Analyst
{Oh please someone help me} OK, let’s say if I was to give you the whole dump of millions of rows of data in the data warehouse.
How would you like to receive that?
Veruca Salt
Now I wouldn’t be the most technical but sure just stick it into Excel and email it over.
One of the other guys here knows how to do pivot tables and we can have a poke about through it and see what we find.
(Screen fades as our heroic Data Analyst bangs their head off the desk repeatedly until the warm embrace of unconsciousness rescues them and snuggles them to it’s homely bosom.)
Sound at all familiar?
Their thought process is obvious.
Step 1: Get all of the data.
Step 2: ….?
Step 3: Profit.
Bona fide 100% guaranteed to successfully nail the metrics you need to push that project into Green status. The big win that cements their future career progression as the brightest young thing of the project management world.
Isn’t it?
No. Of course it isn’t.
So why do they do it then?
It's the lazy option of not having to think about what they are trying to solve and it’s frustrating as hell.
It also really highlights the importance of building a robust process for requirements gathering.
The data team has to be involved as early as possible in projects. It's no good just being a mailbox waiting on a cold request arriving from the business and blindly being delivered without further question or input.
I get embarrassed for the person asking this question. The simple fact is that they are out of their depth but don’t necessarily want to admit it, either to the data analysts or to their managers.
Or worst of all, to themselves.
What if they are just hedging their bets?
I’ve had customers who say they would rather ask for something and not need it than need it and not have it. This is another issue altogether.
Running a small analyst team means we only want to strip requests down to the bare bones of what is really needed.
What this actually is might not always be easily apparent. But it’s another important skill to level up on for the requirements gathering and general business education stage.
Leading From The Front
As analytics becomes more vital to the business, it becomes even more necessary to take the lead and set the agenda.
If you have any plans for progressing your career as an analyst, you will have to embrace the pain of sometimes telling customers what it is they REALLY want. Not what they THINK they want.
Rob Collie and Avichal Singh put the conundrum perfectly in their book “Power Pivot and Power BI: The Excel User’s Guide to DAX”:
“Human beings do not know what they need until they have seen what they asked for”
This x 1,000,000.
Part of your role as an analyst is helping them visualise where they want to go and what they want to see when they get there.
If you only want to do the technical work then that’s fine. There are more than enough tech-only paths to take in analytics. Just remember that you lose the opportunity to craft the direction of the requests if you remove yourself from the customer facing part of the process altogether.
So how do we handle the Veruca Salts without dumping them down a garbage chute in the data factory?
I’m thinking back to similar requests over the years and shuddering at the memories.
Part of that comes from remembering my own unwillingness back then to assert my own knowledge on older, more experienced stakeholders. There was a certain reluctance to not just give them whatever they asked for.
Sometimes it takes time to learn and grow the confidence to do that. Having good air cover from your manager helps.
The frustration of getting in a request like this never goes away. These days I treat it as a puzzle that needs solved. An investigation that needs a little creativity to tease out the real underlying question. Sometimes that makes it a little easier to deal with.
Sometimes...
(And sometimes I just want to leave it to a crowd of oversized squirrels to drag the offender off to a big garbage chute and let the data version of the world of Wonka do it’s magic. But I’m not really allowed to say that out loud of course.)
Top comments (3)
3 years later just as relevant.
A lot of the time I find users have an issue and a theory as to what is going wrong and ask for a report to confirm this. You spend days making it pretty because you have standards. The report confirms or denies their theory and they ask for some revisions. This can take several rounds of back and forth through formal service desk requests. They then ask for the report weekly to validate the changes they implement have the desired effect but never ask for the report to be cancelled. The whole process can take months of back and forth and weeks of Dev time. If you had just sat with the user and cranked code in the database tools in a pair programming session, not only would they have a greater appreciation of what you do, but you will have saved everyone a great deal of time.
Great post, I can definitely relate to this.
The biggest frustrations are being asked for many (sometimes unrelated) data points just in case they are needed, and being asked for them just as the campaign/big senior leadership report/product launch/deadline is due.
I’ve found scheduling a scoping session helps with building requirements but have struggled with being asked for data at the last second?
Beyond pushing back with ‘the team needs xx days notice as we have a significant backlog’. How do you encourage a culture of giving the analyst team time to create the code, review the code, test it’s doing what they think it needs to be doing?
All sounds very familiar Helen! Over the years I've tried pretty much everything in an effort to ward off those horrendous, last minute, "give me everything right now" requests.
Setting expectations that requests have to come in xx days before they are really needed takes time to bed in but has been useful.
I don't think there is any substitute for building up personal relationships with your major stakeholders though. Having a strong working dynamic, really helping them when you can and being upfront about why something can't be done right away goes a long way to getting some respect for the process and lead times necessary for good outputs.
Obviously every business still has it's assholes who will disrespect you, your team and your process whenever they feel like it but dealing with them is another matter entirely :-D