Big Data Marketing for Small Businesses: Why It’s So Damn Hard

Data-driven marketing is such a buzzword these days that it pains me to admit how much I love it.

I love turning numbers into tactics, and that comes from identifying trends in what the data tells you.

And that’s all well and good when you have legion of database engineers and SQL geniuses to help you leverage that data. But what happens when you work for a small or mid-sized business (SMB)? How can you effectively and efficiently turn your data into revenue?

I recently started working for a SMB in Florida called ReStockIt.com as their eCommerce marketing manager. The company is a business supply partner for things like restaurant, medical, and office supplies. With something like 200,000 SKUs, there’s a LOT of data to manage.

Like many SMBs, we’re limited by our resources: both dollars and employee time. There are only so many hours in a day for our (very talented) engineering and IT teams to work their magic.

As a marketer, that presents a challenge. How can I efficiently manage my time to best deliver on revenue and order goals?

I’ll take it one step further: what happens when you have uncertainty about your data? Maybe you don’t feel comfortable about your analytics, or your internal attribute structure needs improvement. In the case of analytics, how would you know it was broken? If you have 200k SKUs, how do you scale out a systematic attribution fix?

(Disclosure: I fortunately don’t have either of these issues. I have, however, dealt with them in past positions.)

If you work for an SMB, the answer lies in the weeds. I can’t recall how many times I’ve rolled up the sleeves of my white-collar shirt and rooted around in Excel spreadsheets; building pivot tables, manually entering product tag data, or writing the ad framework for DKI-based Adwords campaigns.

Efficiency is key here, and so I simply start at the top.

Pick the top 100 selling SKUs you have, and take the time to roll out the changes that need to be made. Let them be your template. If you’re working product descriptions and tags/attributes for your internal data feed, that means manually going into CSVs and writing the values you need. Or dig into your email provider data and root out customer segments.

Test it. Watch for lifts in whatever KPIs you hold dear to those 100 SKUs, and if it works, move forward.

Don’t outsource it to India. It won’t work. I’ve tried that. You’ll waste six months, you’ll only get half your money back, and you will end up starting back over from scratch anyway.

If you’re lucky to have the resources, hire some interns. They’re not expensive, and they want the experience anyway. Once you set a template with your top 100, have three of them roll that out over your top 5,000.

If you haven’t completely failed in your testing strategy, you now have 5,000 highly performing SKUs/keywords/ads/etc that are driving more revenue, and creating breathing room. Use it.

Build your case to your CEO for the need to change the system. Change how new products enter your database. Change how your agency creates keywords and ad copy. Rebuild your automated email marketing streams. Kill the sacred cows.

It will work, because you spent your time in the weeds, and built the case for change. That’s the joy of data-driven marketing. The numbers are there for you to build your strategies upon.

Is it hard? Yes. Tiresome? Absolutely.

Even mind-numbingly boring sometimes? You bet.

But that’s part of the pride that comes from working with a SMB. You can see the fruit of your efforts. When you pull the lever, you can actually see it reflected in the KPIs.

That kind of satisfaction is harder to find in the mid-level marketing roles at the Fortune 500.

Leave a Reply

Your email address will not be published. Required fields are marked *