In 101 Ways To Win An Election, Ed Maxfield and I wrote that:
Data is your organisational lifeblood … You need to love and cherish data, and keep a close eye on both quality and quantity. Far too often, campaigners fail to keep an eye on either, resulting in lost opportunities and mistakes made.
How do you avoid that trap? Here are my five top tips.
1. Backup, backup, backup
The office wall in one of my former jobs had a cartoon with two drunks slumped in an alleyway bemoaning their fate. One was saying to the other, “It all started to go wrong when I realised the backups hadn’t been working…” He at least had been trying to use backups.
Think how much time and money goes into gathering data. Ask yourself if you and everyone else with access to the data is infallible. Put those two thoughts together and, assuming you haven’t discovered you have a bevvy of cloned Popes on the campaign, it means that there is a good chance that at some point something will go wrong, putting at risk data that has had blood, sweat and tears go into its collection.
You can either be cavalier, decide you don’t care about such measly risks, secretly tell yourself that you are infallible, decide that if anyone else makes a mistake that’s their fault for being a fool, and plough on regardless. Or you can be serious about making backups.
In practice, backups are there as much to protect you against a daft human error as against a horrible IT failure. Even if your email system of choice, for example, keeps running smoothly all through the campaign, you still might wreck things by doing a select all and delete by mistake at a crucial moment.
Because of such risks, sometimes people fear trusting data to computers, worried that a wrong key press may result in valuable information being lost. That is to get things wrong: data is safer on computers because it is much easier to do regular backups. Data stored any other way is difficult to backup; reams of photocopies are no match for the simplicity of a computer backup. If you want your data to be safe, give it to a computer and then do regular, proper backups.
However, even the best computer systems can go wrong. Data disasters can and do strike highly reputable services and when the risks of hacking and human error are thrown in, not to mention the cost of losing data for a critical few days whilst someone else sorts out restoring it, it makes sense to back up your data wherever it is.
So make sure there is a backup of everything – not just voter data but also supplier contact details, budget plans and the like.
Backup here can also mean ‘make a copy of in a different place’. For example, I’m a heavy use of MailChimp and it doesn’t come with an option labelled ‘backup’. But it does come with a download option, so when my diary reminds me to, I download the data and store it safely password protected and encrypted. (For no longer than my GDPR-complaint data retention policy says, obvs.)
Likewise, creating a set of pdfs from Connect of your shuttleworths just before polling day is a good insurance against a mini-multitude of problems, including the possible loss of internet connectivity from your polling day committee room.
There’s more about the importance of backups and how to have a sensible system in Chapter 41 of 101 Ways To Win An Election. As that chapter starts: “You can’t stop things going wrong; you can stop them turning into disasters.”
2. Always, always, always check your data before using it
Back to that blood, sweat and tears. If you’re going to ask volunteers to give up their spare time knocking on doors, delivering letters or making phone calls, don’t they deserve to be treated well? Everyone says yes at this point – but saying yes is easy and glib.
If you really believe that they should be treated well, then you will never run a data export and put it to use to generate emails, letters, lists or labels without in between doing some spot checks to make sure the data you’re about to ask others to give up their time to use is actually correct. It only takes a minute or two. It could save hundreds of hours of badly used volunteer time.
Again, unless you’re the Pope, it’s foolish and rude to skip the checking and just hope that you’ve not made a mistake which might waste the time of your volunteers, eat up hard to raise campaign funds and squander precious votes.
3. Design systems to fail well
Of course, what’s even better than making a mistake and spotting it in time is to make a mistake and have a system that copes well with it in the first place.
The classic example is the default value with merging letters or emails. What does the salutation field default to if it’s blank by mistake? Do you end up producing messages that start “Dear [blank]” or “Dear <error! data=notfound>” or, have you smartly set a default and in such cases of problems it reverts to “Dear friend”?
Default values on merge fields are only a taster of what you can to do design systems well so that they minimise errors and handle mistakes well. Get thinking what else you can do. Assume everything will go wrong at some point and think through the consequences.
4. Pull on every loose string
When you spot something not quite right, it’s often tempting to dismiss it as a temporary glitch. Yet it’s also often a clue to something more significant at work. That one name that shouldn’t have appeared in the print run of target letters could be just one name wrong… or could be your clue to a problem with the data selection that means there are many other names wrong too.
The only way to know is to take each clue that you get seriously. Pull on every loose string, no matter how small it looks, and see what unravels as you tug.
5. If it’s hard, you’re probably doing it wrong
There is an awful lot of hard work to do, so this rule isn’t perfect. But I’ve included it as it is oh so common for someone, with the best of intentions, to face a task, think there’s only a time consuming and laborious way to do it and plunge in, working furiously hard – doing something a really slow, inefficient way because they were so eager to get the job done they didn’t first pause for thought, ask around and discover an easier way of doing things.
Not all hard or long-winded tasks can be cracked, but many can. So always think first if there might be a better way of doing things.
For more about how to get the best out of data, see 101 Ways To Win An Election which includes several chapters on this very subject.