Sherri-Lee Mathers of Balsam Way Bookkeeping knows the value of ensuring that data is not only clean and reliable, but that it best serves the specific needs of each of her clients and their teams. For Sherri-Lee, communication and documentation are the cornerstones of getting clean data into the systems she uses to manage her clients’ numbers.

Within the hospitality niche she serves, Sherri-Lee knows that data isn’t always top-of-mind for her clients. It takes time for them to see value in their numbers and to use their data to inform operational and financial decisions.

“We have to know our clients and the industry they are in,” she says. “Data can be highly personalized for each client in the context of their industry. For instance, data in brewing apps used by craft breweries is very different than the information produced by an e-commerce-based business. A one-size-fits-all approach doesn’t work. Niching helps you ensure that with each tech integration, data is moved from one platform or app to another consistently.”

This kind of personalization helps Sherri-Lee get her clients invested in their data: if they feel the data is relevant to them and that it’s directly helping their business grow, they’re likely to take ownership of keeping data clean and doing regular data hygiene.

 

What is your biggest data obstacle? Where do you find that dirty data tends to live in your client’s records?

Dirty data can hide in many places, it really depends upon the user. Within, say, the sales department, it’s usually issues like incorrect or replicated customers, or wrongly recorded revenue, while operations often have costs assigned to the wrong departmental code, and in the accounting department, they might find dirty data showing up in the balance sheet.

 

What do you see as the main barrier to creating or working with clean data, consistently?

The main barrier to clean data is communication. It’s vital that there’s an understanding of the objectives for data management by all stakeholders. Those involved need to understand how the data can or will be used, who needs the data, what kind of data will be extracted, and when it’s needed. 

Take a craft brewery: the head brewer is always in research and development mode, always creating different brews to satisfy the tastes of the customer. They need data like ingredient costs to be exact to measurement (i.e. 20g of yeast costs x-amount) so they can understand batch costs. If the price of an ingredient goes up dramatically, they need to know how different recipes are affected and how the margin is being affected. This kind of thing can then also affect packaging: is it more cost-effective to package this beer in cans for resale or keep it as a custom batch for your own taproom?

 

How does having clean data help you help your clients?

The average small business owner often isn’t fully aware of the value of having data at their fingertips, never mind understanding the meaning and importance of clean data. What they do understand is having access to information that helps them make informed choices so they can be more responsive to change.  Again, I’ll use the example of a craft brewery: if say, citrus prices dramatically increase, the brewers can see the planned production costs and pivot away from making a citrus brew in favor of a brew that won’t cost as much.

It’s important to know who will use the data and how they’ll use it, and to take that into account when communicating with clients. Someone who isn’t handing data all the time, like a sales staff member, has different needs when it comes to information and how urgently they need it than a “power user” of the business data management systems, like the business owner or controller.

 

What do you do to engage your clients about their data?

High tech equals high touch and greater client engagement is the result! Traditionally, the discussions have been about having consistent data entry so they can see comparatives in their financials. But we now see more data silos––data living in different areas or systems––and apps bringing in different information. So, we’re now asking clients up front about what answers they are regularly looking for in their business data and who in their organization will be asking these questions. This triggers more involvement from stakeholders. For example, when the head brewer understands he can get specific information from his data, he is more inclined to ensure costs are entered to the brewing system correctly.

Education is also key because different users in the system have varying skills or levels of data literacy. Resources and documentation are essential for the business to be able to take ownership of their data. 

 

How does clean data impact decision-making for your clients? 

Up front, you need to know how much the client values their data and what different users expect to get from that information. Blame or cause aside, dirty data leads to mistrust. A business owner is using the data to make decisions, so they must be able to trust that the data is accurate. Think of trusting your mechanic or your doctor: early in the relationship, it may have taken time to build a relationship and view their analysis as totally credible. You have to make sure the professional really understands your individual needs. It takes time to develop your trust in them, and this is often a result of ongoing analysis and whether the recommendations they provide are consistently correct and helpful.

A business owner will develop trust in data when you can prove its quality and usefulness. Imagine querying sales data for the year by region, but a number of the invoices didn’t have the state consistently entered. The data might reflect 10 variations of the same state name. In other words, this data isn’t clean. There isn’t a single source of truth to work from. The client looking at this data sees the sales from that state as under-reported and assumes that there’s an error. They see conflicting information from different sources. Immediately, this causes a lack of trust in the data management software systems, or worse, in your expertise, which can lead to distrust in the data and abandonment of the software.

 

Do you feel that manual data entry is a barrier to creating clean data? Do you think technology and automation can help solve this?

Manual data entry slows down the availability of usable data and opens up the possibility of increased inconsistencies in the data. Technology and automation can speed up data entry, but this brings its own unique challenges. One program may abbreviate a state by 2 letters, another by 3. I see this a lot with differences in date formatting in various systems. There are also differences in what data each system actually uses: one program might bring over a summary journal entry, combining all the sales to one figure, while another sends in each and every sales transaction individually. What’s critical is having a set of documented standards to ensure information is entered the same way across all systems.   

As more and more businesses understand that they have valuable information available to them, they want better, faster access to it, and they expect a self-service approach. The challenge is that most users don’t understand that it can take time time to learn and implement a self-service system if they’ve never had the opportunity to interact with their data in that way. Users expect that self-serve options will solve all of their problems right away. They forget that they have to know the best way to input data — clean data — so that it can be accessed and analyzed later on. Even with extremely intuitive technology, there’s still an element of figuring out how to optimize it for your specific needs.

 

This Q&A is part of our clean data blog series. Next up, Erich Ly shares how he helps his small business clients take advantage of their data to spend more time working on their business instead of in their business. In part one, Anna Ready shared how she optimizes data for her non-profit clients, so they can focus on their cause and not on issues resulting from bad data.