As reported on TechCrunch.
by MIKE GHAFFARY
Marc Andreessen and Reid Hoffman recently debated whethersoftware would eat traditional retail, leaving no brick-and-mortar presence behind. Both articles noted that e-commerce is currently only 5 percent of retail in the U.S., while the other 95 percent is brick and mortar.
While Andreessen holds firm that e-commerce will completely overshadow physical retail, the debate missed why some industries will never go completely online.
Some industries will have a hard time competing as more consumers embrace online shopping, but some industries will do just fine. I have constructed a simple formula that can show that the magnitude of offline local commerce is likely to be larger than e-commerce for a long time. These calculations could have predicted that Amazon was going to do very well in categories like books and iTunes in music, but the $300 billion clothing industry is going to stay mostly brick and mortar for the foreseeable future, as are many service-driven local businesses.
One quantitative way to gain an understanding of which local businesses will be eaten by software and which will live on is to understand the local coefficient. The local coefficient (L) attempts to define how “local” each category of product or service must be, normalized on a scale of 0 to 1. You could even pick a number from 0 to 1 for each category (e.g. 0.2 for books, 0.9 for restaurants, 0.6 for clothing, etc.) based on your own intuition of how “local” it must be, but I tried to make the assessment more granular by breaking it down into three parts.
Here is the equation I use to calculate the local coefficient:
L = (e + t – s + 5) / 15, with each of the following input variables ranging from 1 to 5:
- e = experiencing the service or product in person after buying it is important
- t = trying, touching, or seeing the product or service immediately before buying it is important
- s = substitutes are available online from a reliable source
Table 1 below shows my assessment of the local coefficient for a variety of industries. There is certainly a fair amount of subjectivity in choosing the variables e, t and s. For example, I assumed that for books, the desirability of experiencing them physically (e) is only a 2 out of 5. While some people say they still prefer the touch of paper as they flip through their favorite novels, the trend seems to indicate that the convenience of having an entire digital library of millions of pages at your fingertips in a device as light as an iPad mini or a Kindle is more appealing.
In addition, the need to try out books in person (t) is low at a 2 out of 5, as evidenced by the number of large bookstores shutting down. Finally, Amazon/Kindle and other sources provide an excellent online substitute (an s of 5); contrast this with doctors, for example, where WebMD provides an online substitute that is only a 2 out of 5.
The book industry, then, has a local coefficient of 0.27 — a pretty low value on a scale of 0 to 1. Thus, you would expect the book industry to move online more quickly than other industries, and it is no wonder that Amazon chose books as its primary market to launch with in 1995. It is also natural that music followed, given that music also has a low local coefficient.
Table 1 – Local Coefficient By Industry
| Category |
e |
t |
s |
L |
Largest Online Substitute | Largest Site to Find this Type of Local Business |
| Restaurants |
5 |
3 |
0 |
0.87 |
– | Yelp |
| Apartments |
5 |
2 |
2 |
0.67 |
– | Craigslist |
| Books |
2 |
2 |
5 |
0.27 |
Amazon | Yelp |
| Cars |
2 |
4 |
2 |
0.60 |
eBay Cars | Edmunds.com |
| Groceries |
3 |
3 |
2 |
0.60 |
Safeway.com | Yelp |
| Clothing |
2 |
5 |
2 |
0.67 |
No clear leader | Yelp |
| Shoes |
1 |
5 |
3 |
0.53 |
Zappos | Yelp |
| Hotel |
5 |
3 |
0 |
0.87 |
Priceline, Expedia | TripAdvisor |
| Spa |
5 |
3 |
0 |
0.87 |
– | Yelp |
| Fitness/gym |
5 |
4 |
0 |
0.93 |
– | Yelp |
| Plumber |
5 |
3 |
0 |
0.87 |
– | Yelp |
| Music |
2 |
2 |
5 |
0.27 |
iTunes | – |
| Doctor |
5 |
3 |
2 |
0.73 |
WebMD | Yelp |
| Dentist |
5 |
3 |
0 |
0.87 |
– | Yelp |
| Legal |
2 |
4 |
2 |
0.60 |
LegalZoom | Yelp |
While the assumptions on e, t and s in Table 1 are debatable, the implication in Table 2 is pretty clear that the local market is going to stay larger than the e-commerce and online-only market for some time to come, even if the local market continues to become heavily online-influenced. Table 2 shows local being 3x larger, in fact.
Categories of products or services with a high local coefficient are going to remain as desirable in-person local experiences. We can confirm the results in the table above with intuition. Obviously, you can’t eat food over the Internet, so restaurants will always have a high local coefficient. Dentists, spas, and plumbers are all similarly high, with local coefficients of 0.87; you really need that in-person visit, and you probably want to meet the service provider before you buy. In addition, there really isn’t an e-commerce substitute for any of those categories (although if it could somehow be done, there would certainly be demand for an online dental cleaning, or better yet, an online massage).
Legal services is an example of a category that is somewhere in the middle: Talking to your attorney in person is very desirable for many potential customers, but ultimately the final work product often doesn’t require you to be there in person (unless the attorney does a bad job and you wind up in court!), and there is a decent online service substitute with LegalZoom and related services.
With clothing, there is at least a medium desire to try clothes on in a store, as fit is an important part of the selection process for many shoppers. In addition, there is not a clear online substitute to trying on clothes other than buying them and shipping them back and forth (e.g. Bonobos, Warby Parker and Zappos). Startups like Clothes Horse, Fit Valet, Metail, and Fits.me are all trying to create virtual clothing fitting rooms, but there is still work to be done.
Now we can construct an updated table with market sizes (M) for each category (click links for sources) and estimate the size of the steady-state, local brick-and-mortar industry (L * M) compared with how much will go online (O * M):
Table 2 – Market Sizes
| Category | e | t | s | L | O | Market in US (M) | O * M | L * M |
| Restaurants | 5 | 3 | 0 | 0.87 | 0.13 | $660 B | $86 B | $574 B |
| Apartments | 5 | 2 | 2 | 0.67 | 0.33 | $127 B | $42 B | $85 B |
| Books | 2 | 2 | 5 | 0.27 | 0.73 | $33 B | $24 B | $9 B |
| Cars | 2 | 4 | 2 | 0.60 | 0.40 | $168 B | $67 B | $101 B |
| Groceries | 3 | 3 | 2 | 0.60 | 0.40 | $491 B | $196 B | $295 B |
| Clothing | 2 | 5 | 2 | 0.67 | 0.33 | $305 B | $102 B | $203 B |
| Shoes | 1 | 5 | 3 | 0.53 | 0.47 | $48 B | $22 B | $26 B |
| Hotel | 5 | 3 | 0 | 0.87 | 0.13 | $137 B | $18 B | $119 B |
| Spa | 5 | 3 | 0 | 0.87 | 0.13 | $13 B | $2 B | $11 B |
| Fitness/gym | 5 | 4 | 0 | 0.93 | 0.07 | $21 B | $1 B | $20 B |
| Plumber | 5 | 3 | 0 | 0.87 | 0.13 | $88 B | $12 B | $76 B |
| Music | 2 | 2 | 5 | 0.27 | 0.73 | $7 B | $5 B | $2 B |
| Doctor | 5 | 3 | 2 | 0.73 | 0.27 | $800 B | $213 B | $587 B |
| Dentist | 5 | 3 | 0 | 0.87 | 0.13 | $94 B | $13 B | $81 B |
| Legal | 2 | 4 | 2 | 0.60 | 0.40 | $245 B | $98 B | $147 B |
| Total | $3237 B | $816 B | $2421 B |
The results in the table should be surprising for anyone who believes brick and mortar is dying: Even after we all become compulsive smartphone users who buy anything online whenever possible, the theoretical equilibrium of local commerce versus online commerce will be 3 to 1.
The opportunity for mobile and web companies now is in online-influenced commerce for these high-local-coefficient categories. Online-influenced commerce means that consumers want a mobile or web experience before, during and after going to a local business. Forrester Research sized online-influenced commerce as already over $1 trillion back in 2010.
Let’s take restaurants for example. For many of us, the ideal experience before going to a restaurant is going online to find somewhere great to eat and explore ratings, reviews and menus. Then at the restaurant, we want to look at review highlights to see what to order, as well as photos of dishes. After we leave, we might write a review and rate the business. This kind of online-augmented local experience is becoming so ingrained in our normal behavior that The Onion finds humor in suggesting someone might actually go to a restaurant without reading online reviews first.
You can also see that the categories with the highest local coefficients map nicely to the most popular reviewed categories on Yelp. This is no coincidence and is part of the same trend:
Reviewed Businesses by Category:
- Shopping & Retail – 23%
- Restaurants – 21%
- Home & Local Services – 11%
- Beauty & Fitness – 9%
- Arts, Entertainment & Events – 7%
- Health – 6%
- Nightlife – 4%
- Travel & Hotel – 4%
- Other – 11%
The next wave of innovation will be less about how to deliver an experience online, and more about how to enhance it. While the assumptions behind the local coefficient are subjective and can be debated, even if you make very aggressive assumptions, it is hard to make the calculations show that e-commerce will become larger than local brick and mortar in the foreseeable future.
My analysis resulted in brick and mortar being 3x the size of e-commerce, but you can probably find a way to get it to 2x or even 1.5x. But getting to 0.5x would probably stretch beyond the set of assumptions that a reasonable person would agree with.
Changes in society will occur in the coming decades that could make all the input variables shift, but it is hard to build a mathematical case that e-commerce will surpass local by the end of this decade. Entrepreneurs would be wise to focus on online-influenced commerce and how to create great experiences in that area.