Life-saving data management practices for marketers.
Databases. Every digital marketer that does multi-channel engagement, or has a sophisticated sales & marketing team, will definitely need to deal with one. Doesn't matter if it's Hubspot, Salesforce, Dynamics 365, SAP Business Suite, or any other CRM provider, they all consolidate customer data needed by multiple departments in your company.
This means many fingers in the pie. Stakeholders with varying degrees of IT literacy, contributing non-standardised, incomplete, and/or erroneous sets of data. Colleagues who insist on working off spreadsheets. Without practising good version control, AND data hygiene. (I'm not going to discuss strings vs values, and formatting today. Gotta pick the critical battles, and work with the usual short attention spans.)
If you've had to manage CRMs, you'll understand what the sum total of this looks like. Gaps in the table. Duplicate rows of the same prospect. Duplicate columns, which makes updating client data sort of like fighting a multi-headed fire-breathing dragon. (If you have short-tempered clients, that fire is real, at least for your client-servicing teammates.)
These problems are probably managed by your Data engineers, especially if you work for a large organisation. (They are the real unsung heroes, ok.) But if you don't have that luxury, read on.
Preventative Actions
ONE. Integrate all your databases, and keep a master database.
Many companies use multiple tech solutions. Which is fine, as long as the databases link up with each other.
Integrations with existing tools used by the company should be a key consideration factor for purchase. Minimally, there should be single directional syncing.
There should also be a team member who loops the info back to other databases in your system, and does periodic checks to ensure the information syncs are correct.
Otherwise, your clients can update your accounts receivable team three times, without your sales team ever getting the memo. And that's the most benign problem possible.
Oh, and do maintain a verified master database. This will likely save so many of your colleagues, and yourself.
TWO. Do a quarterly backup of your CRM.
Weekly backups, even, if you are uploading on a frequent, almost-daily basis. This will protect you from catastrophic damage, because you will be able to rollback mistakes. Also, be sure to use a cybersecure, encrypted storage option. (Yup. PDPA obligations.)
THREE. Pace your uploads
If possible, limit uploads to the CRM to once a day. Why? It'll be easier to pinpoint the errors and allow for a roll-back or do-over.
FOUR. Verify all spreadsheets and customer lists before you upload.
Random-check the data at the top, middle, and bottom of your table. It'll protect you from erroneous data that result from skipped rows.
FIVE. Work off the CRM, and get your colleagues to use them too. NOT spreadsheets.
It's 2020. When you use CRMs, you'll get so much more information on the clients at your finger tips.
Which page on your company site did your client visit? Did they attend your previous conference/event?
Did they open the email? (If you use Hubspot, you'll know when they opened your marketing email, on which browser and type of device. Even if they deny it.)
Spreadsheets simply can't help your company level up its game like that.
Of course, there could be various reasons (or stakeholders in your company) that results in the need to use spreadsheets. Here's how you cope, and what your teammates must do.
Limit spreadsheets (that contain client info, or needs to be uploaded to the CRM) to one per team
If you can't have a single source of truth within your department, how are you going to sync the information up with other departments?
Use Google Sheets (or practise good version control)
If you need to work on a spreadsheet simultaneously with other teammates, you should totally use Google sheets. Version history is a godsend.
It's not the end of the world, if you're stuck with non-live-synced, on-machine file editors, like Excel. You'll just have to practise good version control, by documenting revisions and using file naming formats that include dates. (And, of course, good coordination with stakeholders involved.)
NEVER delete columns
Every time you delete a column, a unicorn dies. Here's why.
CRMs usually have features to manage duplicates and property updates. Which usually work by matching email addresses, object IDs, and/or other unique identifiers.
So, whenever you delete a column, your CRM will lose data points that can be used to reconcile information on existing clients in the database.
Also, if you and a member of another department discover that identifier numbers & emails don't match up, you'll both know there has been a mistake in one (or more!) spreadsheets. Which presents an opportunity to spot and rectify the issue.
So, minimally, ALL your client data spreadsheets MUST contain email addresses AND internal object/client IDs.
There may be times where all the above is not possible, and your worst nightmare happens. Wrong information gets uploaded, and a lot of client information gets affected. Here's what you do, to fix the problem in a short time.
5 Steps to Fix a Wrong/Bad Upload:
ONE. Assess the damage.
How bad is it? If the upload did not affect existing clients in the database, simply delete the new contacts (using time-based filters), correct the file, and try again. The End.
TWO. If existing clients are affected, try a roll-back
This is a viable solution only if you have a very recent backup of your database, and any data lost during the rollback can easily be re-added. The End.
THREE. Isolate and fix the affected. (Manually.)
Usually, this is a simple matter of pulling up the offending uploaded file, and correcting the rows with mistakes. Viable to fix manually, up to about perhaps 100 rows. More, if you have supportive teammates. (Read: Others who are competent and jointly responsible for the screw-up.)
FOUR. Isolate and fix the affected. (With data validation and VLOOKUP.)
Ok. So, you have 500+ rows to fix. Take a deep breath and don't panic. Now, go find a reliable list of existing clients that you can use to compare the information with. This is usually an earlier backup, or a dataset from another department that contains email addresses and/or internal object/client IDs. Now, you'll be able to match and correct discrepancies. (If you can't manage this, and you can't bribe a data engineer to save you, well, you're in serious trouble.)
FIVE. Verify and upload the correct information this time.
In general, if issues can be fixed and the persons responsible take corrective action, I suggest refraining from naming-and-shaming. But if you face recalcitrant offenders, either report the action or move on from the organisation. Spending too much time fixing CRM problems will only prevent you from doing marketing efforts that contribute to the bottom line. Hope this helps!
