Your enterprise resource planning (ERP) solution manages information across your business, and you’d be lost without it. But ERP systems are only effective if the data they contain is accurate—all the time. 

Accurate ERP data is paramount, but so many direct materials manufacturers struggle with getting it clean and keeping it clean. 

In June, Sarah Scudder hosted a webinar featuring a panel of experts. Their focus was ERP data—how to get it accurate and keep it clean into the future. You can watch the recording of the webinar here.

Let’s take a look at a few highlights. There are a lot more great insights in the full webinar so give it a watch!

What have you found to be the top causes of bad direct spend data in an ERP?

Peyton Whitehorn, Integration Advisor at Reveal:

As a data nerd, I start at the higher level, which is poor master data governance.

• Lack of approval processes
• Misspellings and duplicates due to unassigned roles and responsibilities
• Lack of uniform data entry processes
• Outdated pricing
• Materials grouped improperly
• Spend categories are inaccurate
• Delivery dates not captured against the actual promise date

Mathiew Pappalardo, Chief Procurement Officer at GoGo Squeeze

At our company, the delivery data is almost always wrong, prices change every quarter, and POs we cut are very fluid.

So the essence of the data is that it’s wrong to begin with, so the question is, “How do you make sure it’s correct and gets updated when it needs to be?” 

Unless you have a tool that does that for you, you need a process. You need a way to have that data constantly updated to reflect reality.

Then the problem becomes, “Who owns the process?” For example, if we book 5,000 tons of a specific puree out of South America, we negotiate a deal and let the plants cut the POs as needed for production. Now, what if the PO is wrong on the price or the delivery dates? Who wants that?

Is it the plan people because they’re cutting the POs? Or is it procurement because they own the relationship with the supplier?

So process is key, but ownership is also key. If you don’t have all of that, then you have bad data.

Susan Walsh, Mistress of Data at The Classification Guru

I’m Susan Walsh, known as the classification Guru Fixer of Dirty Data. I’ve been cleaning, classifying and normalizing suppliers in the procurement space for about 11 years now, so I’ve seen it all pretty much. And I think particularly in the manufacturing space, it’s multiple versions of the same item. That’s a big problem. It’s the same thing, but it’s been abbreviated, slightly worded differently, and it creates multiple versions of the same thing. 

And many times we work with companies and find that their data has been bad for 10 years and is continuously bad. There’s a little bit of laziness, no willingness to actually do what needs to be done. The other problem is that you might have multiple people on your team all inputting that data and they’re all doing it slightly differently. It’s hard to motivate people sometimes to do it properly because it takes extra work and it’s time consuming.

How can you get ownership for master data like inventory items that are entered and managed by multiple teams when no one wants to own the data set?

Juliette Samson, CIO at GenPro:

You must have an item master owner, and it’s important for everybody in the business to know who that is. And as items are getting added, you don’t add it just because you have a PO and you’ve got to do it today, which most people do. And then you add partial information and it messes up your entire workflow because you can’t identify [things like] transportation needs and inventory needs unless you go through the item master process.

We have devised a process called an item data upload form. Anytime new data comes in, we ask the product manager to fill out an item setup form which then gets uploaded to the ERP. The information is not uploaded until all of the parameters are filled in. And you must get sign off from marketing, sales, and finance.

The key aspects of this upload form are size, weight, number of items per case, number of items per pallet, item cost, owner, vendor, and brand. These will help you with logistics so you can figure out things like how much will fit on a truck for example. 

People tend to do this very fast because they have to add this queue and they’ve got to sell it. But it is so super important to keep your data clean and usable, because if part of the data is not there, you can’t transport that SKU, or you’ll get bad data.

Why is it so difficult to get buy-in from key stakeholders?

Mathieu Pappalardo

Ownership is key. If nobody owns it, well, nobody’s going to push for a resolution. If it’s strictly in one department you can be sure that the head of that department is going to push for a solution. If it’s in between departments, who is going to push for that? Everybody is going to get frustrated, but nobody is going to actually act on it.

Also if you want to invest in a solution to fix your bad data, how do you tie that to your bottom or top line in actual financial numbers? Yeah it’s going to improve collaboration, reduce frustration, and improve productivity, but what’s the dollar amount? It’s hard to estimate the return on investment, so for that reason I think it’s hard to get the buy-in unless you have a big visible issue due to bad data and people are ready to act on it.

What are the real business costs of bad data?

Peyton Whitehorn

Yes we see the negative impact on the bottom line, with customer orders being late or not in full. But we also see it in the life of the buyer. It’s greatly impacted by the fact that the data is bad. Spending time on manual processes, answering phone calls because something is late—that’s a huge deal. It creates frustration, lack of collaboration, and demotivation. 

What are some actionable steps for cleaning up bad data and keeping it clean?

Susan Walsh

Whatever you do, it has to be simple because quite often people who are inputting the data are not necessarily data people or have anything to do with that space.

So you need to make it as easy for them as possible. Don’t give them too many options like dropdown menus or free text; this can be a nightmare. It’s important to understand the whole process, share with the people who input the data, and stress the ways it makes an impact for them as well.

How can someone clean up their data so that it better helps with inventory predictions and managing inventory levels?

Peyton Whitehorn

What are the average inventory values? What is the dead stock value? What safety stock are you carrying? Generally, safety stock becomes dead stock, but it’s necessary in some ways. In manufacturing, we started with the Lean methodology before 2020, and then the supply chain changed and everybody started overstocking because they’re afraid of running out of supply.

So right now there’s a major dollar impact on carrying higher average inventory value and ultimately ending up with dead stock. This ties up working capital and it’s a liability for the company. So we want to see those average inventory values go down by cleaning up the master data and ordering what we need when we need it—and no more than we actually need.

Look at the production planning processes, look at lot sizing supplier constraints, update the MRP and lot size procedures to make sure everything is aligning with the production plan, sales plan, etc. Ultimately, we want those available-to-promise numbers to be right and have the right lead times. 

SourceDay keeps your ERP data accurate, in real-time. See how today – get in touch with us to schedule a demo.