The latest ENR 500 came out on April 15th so it’s time we said something about it. For better or for worse, there isn’t much to say. We re-ran the last analysis looking for major changes. The largest gains were seen by companies in the $1-$2.5B segment with a total growth of 40%. The second largest change in the market was followed by firms in the $25-$50MM segment with a total of $14.6% growth. The market overall grew approximately 3.5% in 2013 which is much better than the 1.9% growth in US GDP for 2013. The small overall growth percentage may be disappointing to some. It is characteristic however of a steady growth in the market that is likely to be more dependable then the erratic ~12% growth seen in 2008 that quickly deflated.

ENR Top 500 Combined Revenue

Last week we looked at the ENR 500 and tried to pull out whatever interesting information we could find. This week we turn back to the ENR 500 and take a bit of a deeper dive. We aren’t taking the drawn out, step-by-step approach we did last time. Instead we’ll present some information and add our own insights where appropriate.

It has been about a week since the first post. We hope to maintain that kind of frequency. We will post whenever we have new or interesting information. Our goal with this blog is to aim for quality while providing some kind of regularity. We also want to give enough time to digest what may be “longer than average” blog posts.

In order to look at how the ENR 500 has changed over time, we took the last ten years of data and carved it up to see what might be of interest. What does it look like if we divide the firms into various “income” brackets based on their domestic (US) revenue? Here is that data in table form.

ENR Top 500 Revenue Distribution

“What the heck am I looking at?” If we look in the top left corner we see that one firm earned between $2.5 and $5 billion dollars in 2002. Each value in the table is the number of firms that occupy that “income bracket”. Feel free to draw your own conclusions from these values. Tables are generally worse than graphs at showing changes in time so let’s graph it.

ENR Top 500 Revenue Distribution

This is hopefully a little more interesting. Perhaps a bit dated but we see a very clear (albeit delayed) indication of the recession. Just prior to the economy’s change, most ENR 500 firms were earning over $25MM (again, domestic revenue only). Leading up to the recession we see an expansion in both the $25-$50MM and $50-$100MM segments. This suggests that many firms crossed the $50MM threshold in the years following 2003. That trend started reversing in 2007 and didn’t actually bottom out until 2011. Perhaps more relevant and provocative is the apparent repeat of that shift in more recent years. Firms in the sub-$25MM range are decreasing and the number of $25-$50MM firms appears to be increasing.

This is good news for Consultants and Engineers that want to work for these companies. Perhaps a slight cause for concern however is the apparent lack of growth in the $50-$100MM and $100-$250MM segments. It is difficult to prove exactly what is causing this. We can however look into the possibility that the large companies at the top of the list are reducing the pool of revenue for everyone else. This graph is not the best way to answer that, so we should shift our focus away from the number of firms and more towards the combined revenue of these segments.

ENR Top 500 Combined Revenue

How did we get this graph? We started by taking the revenue from firms that earned less than $25MM, and added them up. Then we took the revenue from firms that made between $25MM and $50MM and added them up. We’re still using the “income bracket” analogy. This time however we combined the revenue for each of the companies in those brackets, instead of just adding the number of companies.

Why is this useful? It gives us one way of seeing where money goes, it also let’s us answer the question about big firms crowding out smaller firms. Prior to 2006, the $2.5B+ segment made up a fairly small percentage of the overall revenue earned by these firms (just under 6%). There was only one firm who even entered this category so not really all that “powerful”. Since then that segment has grown both in revenue and in number of firms. In 2012 there were 3 companies who occupied this segment taking up 17.6% of the market. Certainly an improvement over their position 10 years ago.

Is this a bad thing? It is and it isn’t depending on where you’re sitting. It’s bad if you’re sitting at an A/E/C firm and you’re trying to compete at the top of the market. Higher revenues and more competition means higher risk. It’s also tough for companies sitting near the edge of this top-tier because it’s hard to overcome the momentum of those giants.

The flip side of this coin is the revenue stability in the smaller segments. I refer to it as being “stable” because we don’t see any major expansions or contractions between $25MM and $500MM. This suggests a fairly dependable revenue stream for a certain segment of the market. If this were not the case we might see the lower revenue segments shrinking as they got “crowded out” by the larger firms. According to this data, the market has actually grown to accommodate the new giants and left the other players relatively unaffected (of course, this excludes firms outside the ENR 500).

You may be thinking about this dataset and wondering if we can get a look at what happened with each individual company. We certainly have the data for it. When we tried to graph the individual companies we got a warning in Excel that “you can only graph 255 entries”. Even after pulling out 623 “non-architectural” and data limited companies (firms that had less than 7 years of data) we knew it would still be painful on the eyes. Instead we opted for the interactive visualization below. The graph shows the annual revenue for 200 firms practicing architecture that also participated in the ENR 500 for 7 years or more). We included some regional filters. It’s interesting to get a rough sense of how income is distributed regionally. It may be of interest to note that even though our dataset is limited to 500 companies annually, over a 10 year span we have actually captured data for 845 unique companies (after adjusting for mergers and name changes)

Lastly please note that this is plotted on a log scale to minimize the differences between market segments.

Circle radius = sqrt(revenue/10)

The short answer is, they make a lot of money. It’s going to take a bit to show you how we figured that out. Welcome to the first of many blog posts by Data Cornerstone. The underlying goal of this company is to combine and carve up large, cumbersome, and disparate sources of data into bite size chunks. In time we will acquire, develop and make use of new and newly interesting stores of data. For now we will take a look at more commonly available data to show you why we think Kansas City is interesting and what we are all about in the process.

We’re going to start by looking at the Engineering News-Record (ENR) Top 500 list. An understandable first reaction when one looks at the ENR Top 500 is “Wow, look at how far off the meter those top 8 entries are?”

Top 15 ENR 500 Firms by Revenue

8 out of 500 is an impressively small number of firms for the lead they have on their competition. Collectively the top 8 companies are currently responsible for nearly 40% of the annual revenue on the entire list in 2013 (roughly $35 out of $90 billion). It is a wonder that they have any competition at all. Many would say they are in a class unto themselves.

The ENR Top 500 is one representation of the Architecture, Engineering, and Construction Industry. If we stop there we get a taste for the AEC industry that is overwhelmed by a few powerful but ultimately uninteresting flavors. The headline probably clued you in that there is something more interesting hiding in this data.

When we begin a data analysis project, it is not all that different from cooking. We frequently have to “prune” the data and pull out dirty/unusable bits that might spoil the results. Anyone who has cooked a potato knows you have to peel it, chop it, boil it, and possibly even mash it to make it edible. Such is the process of data analysis which is what Data Cornerstone is all about.

We could start simply by removing the top 8 entries and carrying on. It would be completely valid even from a statistical perspective. These entries are almost definitely what you might call outliers. It’s statistically valid as long as we don’t pretend to represent the entire industry from that point forward (even now, we’re only looking at 500 firms). What we are now considering is a subset, one that will hopefully become more interesting.

I am going to suggest we take one more “pruning” step before we move on. The ENR Top 500 is composed of what is actually a large variety of companies. Ones we may not even be interested in. Architects, Engineers, Environmentals, Geotechnical Engineers, Landscape Architects, Planners and “Other” comprise the entire list of options. A number of hybrids also occupy the list, as in “AEP” (Architect, Engineer, Planner) or “EAC” (Engineer, Architect, Contractor) indicating that the annual revenue at these companies comes from a variety of sources that we may not be interested in.

For that reason I am going to suggest for the intended audience that we also filter out any entries that do not have an “A” associated with their “type”. As an example, we will filter out Geotechnical Engineers but leave in Engineer/Architect/Contractor types. The assumption here is that we are primarily interested in Architects or related firms who might be interested in hiring Consultants and Engineers.

Now that we’ve trimmed the fat from our data, we see that we still have a number of “outliers” but possibly fewer than we had before.

Top 10 ENR 500 Firms
(Offering architecture services)

At this point the differences in revenue are extremely exaggerated. You can almost tell without the numbers that CH2M HILL is twice as big as Parsons Brinckerhoff. If what we’re looking for is a point where we can filter out big companies, we might want to look further down around the $1 billion level.

In this case, it doesn’t matter too much where we cut certain outliers as long as we have a good reason. This one has a good ring to it. We can now follow any results we come up with by saying “for architecture firms with less than $1 billion in revenue”.

Top 15 ENR 500 Firms below $1B
(Offering architecture services)

At this point I’m going to pause and offer an explanation of how this started and where it’s going. I initially wanted to give a taste of what data analysis is like. Even if we started out with a specific question, we don’t want to be closed minded about other interesting things we find along the way. It’s like stepping into your kitchen knowing that you want to cook a steak without knowing exactly how you’re going to make it or what you’ll have on the side

Before I came up with this post I wondered: “Are there any locations specific to this industry (excluding the major cities) that have an interesting amount of architectural revenue?” The answer as it so conveniently happens is “yes”. We don’t even have to filter out the major cities to find it.

Total Architectural Revenue by City
Circle radius = sqrt(revenue/10)

Right there smack in the middle of the US is big ole Kansas City with over $2.2 billion in revenue. They actually have the highest total revenue of all cities represented in the ENR Top 500 when our filters are applied. To drive the point home I will say it another way; from the ENR 500, Kansas City has more total revenue from companies that practice architecture earning less than $1B than New York, San Francisco, Los Angeles, Chicago, and Philadelphia.

“Ah ha!” you might say “Could this be a sign of another outlier?” Sure. It all depends on what we’re trying to find out. When we look back at the underlying entries for KC we see that there are 5 companies who make up the list.

Firm	Type	2012 Rev. ($MM)
BURNS & MCDONNELL	EAC	$933
HNTB COS.	EA	$912
TRANSYSTEMS CORP.	EA	$226
POPULOUS	A	$84
CRB	EA	$79

We might recognize the top two in the table from the results of the filter we applied. We could go even further by filtering them out again and seeing what’s left. While Kansas City is seemingly monopolized by two major firms, it should still be a point of interest. Intuitively many people expect New York and/or San Francisco to stand out. Perhaps this is based on the quantity of firms that we think practice architecture in those cities. Perhaps this even points out deficiencies in using the ENR 500 as a dataset which is worthy of further discussion elsewhere.

I’m going to stop here because I think it’s a good note to end on. We filtered out the companies that weren’t relevant along with the bigger companies on the list. Lastly we mapped the remaining results to see if we could find anything of geographic interest.

I wouldn’t advise anyone to open up an office in KC if you did not have one there already. These results are unfortunately tied to the headquarters of companies that have offices in multiple locations. The data also includes international revenue at these companies which makes it even more questionable. If only we had better data!

Despite all that, you still might want to think twice about the companies that show up on the list above and how your organization is or isn’t serving them. I hope you found this interesting. The detailed lists for New York and San Francisco (the next two respectively on the list of top cities) are below for comparison.

ENR Top 500, Architects in New York City less than $1B