Since unveiling the Unified Data Platform at NADA, I’ve gotten to talk directly to more people in automotive than any time since I’ve joined Foureyes. Hearing from CEOs, CTOs, CMOs, VPs of Sales, and VPs of Marketing about their data challenges and processes, there’s one software product mentioned more consistently than any other: Microsoft Excel.
Maybe I shouldn’t be surprised by that; Excel is a powerful tool for any business. It’s simple to start with. Pretty much every business leader is at least comfortable in it. It’s flexible to handle a variety of needs.
Nonetheless, I am surprised that really significant and sophisticated automotive groups are using Excel to access their data, uncover insights, and track progress.
Why Are Automotive Groups Still Relying on Excel?
My hunch is that this is due to the federated nature of many group relationships. Each store is given decision-making authority around their tech stack, vendors, and marketing budgets. So that leaves groups with, say, 20 stores, 3 different CRMs, and 14 different strategies being executed by 7 different vendors.
Excel becomes a very attractive and approachable place to start wrangling the data to answer questions like:
- Where am I getting my leads?
- How am I closing them?
This becomes problematic when you push Excel too far. Excel was never designed to be a database. And groups have a TON of data. Add these two factors together, and the solution you implemented to get to answers may be leading to false conclusions or have gaping visibility holes that you can’t even see.
From conversations, I’m seeing 7 common symptoms to indicate that it’s time to get your data visibility and reporting out of Excel and into something better equipped to handle a large volume of data.
Symptom 1: Reliance on a “mad scientist”
In case you are your group’s mad scientist, know that I use the term with the utmost affection and respect. I’ve been this person myself:
- 2 AM Thursday: Awake because of a nagging business problem
- 4 AM: Give up on sleep, get up, and open Excel
- 10 AM: 3 tabs, 17 hidden columns, a pivot table, and a graph that’s moving towards an insight
- 2 PM: Share with a few team members. They can’t understand the spreadsheet, but they get the graph. They like the graph. You keep improving it, and that view becomes a staple of monthly reporting.
Or you’re NOT the mad scientist, but you have one on staff. Their graph is good, but you’re worried:
- Maybe they’re biased.
- Maybe they’re leaving the organization.
- Maybe they’re spending all their time in this spreadsheet.
Whether you or someone else is the mad scientist, relying on an individual isn’t a good sign.
Symptom 2: Multiple people maintaining a spreadsheet
Excel has only gotten better at allowing people to collaborate in a single document, and reliance on a single individual has its drawbacks. So multiple people *shouldn’t* be an issue. And yet, multiple people maintaining a spreadsheet is a symptom of a problem because Excel is not designed to be a database. When you start using Excel as a database and invite multiple people to manage it, you open yourself up for errors.
Some common errors that you start to see:
- Row count limitations
- Duplicate records
- Empty fields
- Formatting that conveys information
Symptom 3: A really costly mistake
Data errors like the ones outlined can sound trivial, like minor annoyances that really precise people get hung up on. But once you start using data to make decisions, data errors can lead to costly mistakes. I stumbled on this page that collects spreadsheet mistakes that hit the news. Lost business opportunities. Paying wrong salaries. Inaccurate electoral votes. It shows the variety and relative frequency of active mistakes coming from pushing Excel beyond its intended scope. But I think the pernicious problems are the ones that escape us because we don’t have clean data. For example, how much money do you think the average automotive group spends unnecessarily because of an inability to answer where they are getting their leads?
Symptom 4: Excel running slow
More than any of the other symptoms, Microsoft running slow is the most objective and easy to spot sign that you’re relying too heavily on Excel.
Symptom 5: Data access limited to a developer
A lot of groups have told me about their efforts to get data out of their CRMs. In addition to political hurdles when CRM companies exert ownership of the data input by dealerships into their own CRM instances, there are technical hurdles as well. Many CRMs use APIs or file transfers that are really only accessible to people with coding skills. I talked with one group who negotiated for over a year to get their data from the CRM vendor only to be provided with an API that required a developer’s skills to use--and when they got a developer hired, they found that the API only had a portion of the data they needed.
While this isn’t a failing of Excel directly, it is indicative of how Excel is too simple of a software to match the complexity of the data challenge facing automotive groups. If the data is too hard for a typical businessperson to access, then the tools of the typical businessperson are unlikely to be the right tools for the task.
Symptom 6: Three or more sources
To combine data successfully, you have to normalize the data so it matches up. Take for example something as simple as a lead generator’s name, like TrueCar. Or is it True Car? How about TrueCar.com? Or True car? In our work normalizing automotive data, we’ve seen more than 300 variations of this one lead source in a single group’s dataset.
Additionally, you need to understand the data and the decisions being made to get to the numbers. Take something as simple as open rate on an email. It’s easy to assume that every email tool uses the same formula, but they don’t. Here are three variations I’ve encountered:
- Total unique opens/total sends
- Total unique opens/(total sends - total bounces)
- (Total unique opens - opens triggered by system caching)/(total sends - total bounces).
All three formulas are reasonable, but If you have two vendors using two different formulas, you are going to need to get more data to normalize the open rate.
Symptom 7: Lack of clarity
Maybe the most meaningful symptom of all is that you feel unanchored to your group’s performance. You’re consistently frustrated that you can’t quickly see what’s making one store excellent while another is average, and getting apples-to-apples data feels elusive.
If that’s where you’re finding yourself, know that you’re not alone. I’ve talked to massive groups, really smart groups, and some of the highest-performing groups in the industry and all of them are feeling this pain.
Can Foureyes Help?
Foureyes has spent the last five years wrangling automotive data. Our developers integrated with all the CRMs, engineered a simple script to track all forms, calls, and chats on your website, and collected all the franchised inventory on a daily basis. Our data scientists have explored, normalized, and visualized the data. Now we’re making all of that data accessible to groups through the Unified Data Platform. My hunch is that it can automate and scale a lot of what you may be managing in Excel. Just get in touch. Worried about getting our schedules to line up? Feel free to check out this sneak peek into the tool below.