Businesses face a variety of challenges when their customer address data is poor quality. But what can you do about it even if you know it’s a problem?
Some of the warning signs may be a high percentage of unrecoverable debt, for example. Or slow decision-making and poor business decisions that result in higher-than-expected expenses, losses, or worse.
Maybe marketing and re-marketing achieve poor results. Or customer acquisition and retention costs are higher than expected.
You could experience operational costs higher than average if using customer addresses for service or product delivery. Franchises could experience low returns for their franchisee customers.
There are a lot of ways that low-quality customer address data disrupts business in the worst kind of ways.
What causes poor quality addresses?
Poor data quality isn’t new. Mergers and acquisitions are a big culprit that cause poor data quality. Or the street address capture process could need attention.
For example, if allowing people to enter their addresses as free text, the company almost certainly has low-quality data that costs the business money. South Africa has 11 official languages and a low literacy rate − two attributes that lead to low-quality data when coupled with free text entry processes.
Another major culprit is using data freely available off the internet. Aggregated data seems like a good idea. It’s usually free, or at least very cheap. But it’s unverified and makes assumptions, like assuming all even numbers appear on one side of the road, uneven on the other.
There are a lot of ways that low-quality customer address data disrupts business in the worst kind of ways.
Or it fills in the blanks. Query a street number that doesn’t exist, say it’s number 999 Church Street, and a system that uses aggregated data may drop a location pin at the end of the street without raising a flag because it hasn’t verified that the street only goes to number 65.
The problem is, by the time the delivery agent verifies that, it has already cost money. Or a debt can’t be recovered because a person simply supplied a false address that couldn’t be verified at the time.
Quality data is like a designer wool coat or a pair of good shoes. Look after them and they’ll outlast you. Neglect them and you’ll quickly be bare foot and chilly when you can least afford it.
How you improve address data quality
The best way to make sure it’s right is to make sure to capture it correctly at source. That’s the best time to flag discrepancies and sound the alarm.
It’s also absolutely critical to keep it clean. How clean depends on what the company is trying to do. But, regardless, once the data quality has been improved, you have to keep it that way. You may think street addresses never change but you’d be wrong.
As areas formalise, so the addressing systems change to reflect that. Farms may become extensions of a suburb. Extensions change to reflect the number of developments. Residential addresses fall by the wayside when commercial developers buy up stands and build shopping malls.
The snag is that the original addresses never disappear. No address ever does. They accumulate and compound and that can make the whole shebang a proper nightmare for a warehouse picker, packer, despatcher and delivery driver just trying to get a product out to a customer they’ve never met and probably never will.
That’s why it’s so important that each address has a unique identifier. It’s one of the ways we can track the changes in the data, linking different address formats to the same place on the surface of the Earth, to get a verified and validated and operationally useful reflection of reality on the ground. A unique identifier that businesses can use to connect their products, services and any of their people with their customers.
Applying checks and balances
There are checks and balances that need to be incorporated into address data quality processes.
The first one is easy: ensuring addresses contain all the right fields. Missing fields lead to poor customer experiences. The more missing fields, the poorer the service delivered.
But it quickly becomes complex. Correct entries are a huge problem for businesses that use free text entry. Take Parktown in Johannesburg, for example. It has streets and roads named after numbers. You may need to locate the intersection of 6th Street, 7th Avenue and 1st Avenue West. That’s literally an intersection that exists with a few shops and some residential properties. Even if you get it right, it’s a confusing enough job for a harassed courier to quickly locate.
That’s obviously a worst-case scenario. But imagine missing the town field for number 12 Church Street. How many Church Streets do you think South Africa has? There’s probably one in every second dorpie. You may never figure out where yours is.
The problem with address data quality is that there is no grey area. You can either deliver the product, service, determine your risk, comply or make a business decision that makes a profit. Or you cannot.
Then again, even with all fields completed, the quality of the entry may be flagged. How many ways are there to spell Sean, Shaun, or Shawn Street, Mohamed, Mohammed, Muhammad, Muhamad, or Mouhamed Avenue? Or Blairgowrie? Or Fontainebleau?
We need more than just fields that contain text. We need them to be correct. That’s where autocomplete from a database of verified and validated addresses comes in.
Getting good-quality addresses can be difficult. But it’s the foundation upon which to serve customers. You don’t want to be like the South African council worker who got the location wrong, which ended up in water pipes attempting to defy gravity by having water flow freely uphill.
You need to decide what’s important to you and the business and set up automated data quality processes to make sure this can be achieved. No less and no more.
Share