A Survey of Austin Businesses

How much do you know about your city’s economy?

I came across some data recently that was begging to help tell the story of what the deal with Austin really is…. Austin’s got a reputation for tech and for fun. The tech part isn’t obvious to visitors (and even then I’d argue it also isn’t super obvious to people living here who have a baseline curiosity about tech in Austin — see future posts about that), but with a new restaurant expected to pop up on average every 12 days in 2016, the fun, consumer-facing side of Austin’s business mix has a much bigger share of the economic pie in my mind than it does in reality.

If you experience certain things or hear about certain things significantly more often than others in a set that have a higher actual representation, you’ll likely attribute more importance to whatever those high-resonance things are than what they might deserve. Sorry for the vague language, but this applies to every type of information. To make smart decisions, we should try to consider enough information from the set to reduce that bias, right?

I’ll acknowledge two super important things right now: 1) The data I have is a snapshot in time (August 2016), so it doesn’t give us any information about the change in Austin’s business mix over time. There are boatloads of easily accessible articles about how Austin is changing, but historical data from my source doesn’t have the same search criteria I used to gather the set used for 2016, so, yeah. I agree it’s better to latch your wagon to a rising phoenix than a sinking ship, but even repeating this exercise on historical data wouldn’t paint an appropriate picture of how to find the growth industries. 2) I only pulled data for Austin. So, this is not a comparison of Austin’s business mix versus other cities or the country at large.

What this data does show is a comparison of industry mix in Austin in August 2016 across three dimensions: share of total businesses, share of revenue, and share of employee count. (Side note, there are just over 9,000 “businesses” in the filtered set, which excludes very small and home based businesses and businesses with no website. If you’re curious, the total set of Austin businesses before adding those filters is around 93,000, and around 2,500 of those are government offices.) (The total number of employees employed by the set of ~9000 represented companies is just over 328,000, and those companies reported generating nearly $87B in revenue through the most recent publicly available information.)

The purpose of finding this information is ultimately to help a set of people with general management skills find jobs, not to do a proper economic analysis, so leaving out the home-based businesses, very small businesses, and businesses who are able to operate without a website obviously affects the economic picture, but not materially for the target group. Letting you know in case that knowledge materially affects your acceptance of the data.

First surprise: the wholesale trade makes up 28% of the total revenue contributed by this set of companies. What the hell! Have you ever met a wholesale employee? I haven’t. (And I even know some forestry people.) Maybe that’s because the wholesale trade only employs 5% of the total number of employees. So if you’re looking at “lifestyle” businesses, maybe wholesale’s where it’s at.

Surprise 2: For all that B2B product sitting around, you’d think the transportation contribution would be bigger, but it’s a pretty flat representation of 1% of the total businesses, people, and revenue in the sample! (The industry categories I’ve used come from NAICS/North American Industry Classification System. Transportation in NAICS includes the movement of everything from people to bulk liquids in any vehicle or format from point a to point b, and because of the high capital expenditure and heavy equipment involved, I can’t imagine transportation is disproportionately represented in the super small and home-based business cutoff. Maybe they’re the ones operating without websites. “Do you work in transportation?” “WHO’S ASKING.”)

Nugget 3: If revenue/employee count is proportional to salary (which, categorially, it may not be, especially for high capital outlay industries, but just for fun lets see if it matches intuition), the top industries would be 1) Wholesale Trade, 2) Management of Companies, 3) Raw Materials Manufacturing, 4) Consumer Products Manufacturing, and 5) Industrial Products Manufacturing. The bottom industries would be 1) Educational Services, 2) Accommodation and Food Services, 3) Consumer Goods Transportation and Warehousing, 4) Other (which includes a lot of Repair/Maintenance and Personal Care service industries), and 5) Administrative and Support and Waste Management and Remediation Services. There is a real, better way to figure this out, which I may reveal in another post, but for now, thoughts?

One last comment ‘bout this all…

I’ve been doing a lot of research lately, and one thing has made itself abundantly clear. There are good and bad techniques for understanding data. You can use a bad technique on good data, a good technique on bad data, and a bad technique on bad data, and none of your insights will be good.

And most of the information you find when you Google is an insight, not data. So, thank you to the good organizations, people, and machines who authoritatively and sincerely manage data on things we care about. And especially to those who share for free!