The economic health of the value chain

David ClarkMIT CFP

November, 2012

A more specific question:

• What is the incentive for ISPs to continue to invest in upgrading their service offerings?

• Is there a chance that the cycle of expansion and innovation might stall?

• A hot and very contentious topic– Look at the discussion of my last blog.– Divergence of views at the basic level.

What is the problem?

• (Assuming that there is one…)• Those who use the Internet (either as app

developers or end-users) have benefitted from a continuous improvement in capacity, peak rate, and penetration of access over the last 10 years.

• There is a strong consensus that these sorts of improvements should continue.

• What is the motivation that will drive investment?

Is there anything we agree on?

• ISPs make money.– Both wireline and wireless.

• Loads on the Internet are increasing, largely due to video content.

• Wireline access networks work pretty well today, wireless is less satisfactory– At least in the U.S. Wireless works better elsewhere.

• U.S. regulatory policy is in disarray. – Highly variable situation across the globe.

Next? More dangerous…

• Justification for capital upgrades are hard to make to investors. – Insert hand-wave argument that ROI is about

zero, at least for wireline.• Market in many parts of the world is starting

to saturate.– Harder to justify investment than when one

motivation was supporting new customers.

Where there is deep disagreement

• What is the cost of usage?– Do upgrades actually represent a material cost, or

should we expect that ISPs will naturally upgrade as part of normal cost-reduction?• Obvious answer is that it depends on the rate of upgrades.

• For wireless, are we managing spectrum well?– Are we using spectrum in ways that are:

• Efficient?• Pro-innovation or pro-incumbent?

– Is there a spectrum scarcity?

Going forward

• What is the business model for an ISP?– Triple play? But if everything is over the top?– Agent for license revenues?– Seller of enhanced (managed) services?– Seller of commodity, cost-reduced Internet

access?• Are wireline and wireless the same?– Same business?– Same constraints?

My “follow the money” approach

• What are the sources of revenues that could flow into ISPs?– The consumer (who pays an average of

$45/month today in the US.)• What will these look like in the future?

– Government subsidies and incentives.– Payments from the “other side”:• The sites with which the consumer interacts.• Content, e-commerce, “browsing the web”.

– P2P, etc. does not fit this model.

Who pays for the “other side”?

• The consumer.– Payment for content: NetFlix, NYT, iTunes, etc.• Some content is purchased cross-channel.

– E-commerce. • About $202B in 2011, or $240/month /household

• Advertisers – Online: about $17.4B in 2H 2012• $38.60/month/household

– TV: $75B in 2011.

Conclusions: adverts

• All the “free content” on the Internet is produced and hosted for less than we pay for access. – Or did I miss another source of funds?

• If ISPs tried to tap into this “other side” money, there is not much there.

Conclusions: consumer-pays

• Sellers in high margin businesses might be willing to pay for better QoS.– Example: priority traffic on wireless networks.• 1% of gross wold be $2.40/month /household

• Providers of high-value content might willing to pay for better delivery.– Example: zero-rating (exempt from monthly

quotas)

What is the difference?

• Between direct payment by the consumer and indirect payment via the “other side”?– In some cases, indirect payment may be easier for

consumer to understand.– Indirect payment forces an ISP to set a price point for

enhanced service. • Hypothesis: high variation in value in different

circumstances.

– If consumer pays, will pay for enhancement of what is of highest value to specific consumer.• Analog to bundling of content

Engage the consumer

• Sell enhancement to high-value applications.– Selling QoS

• Charge for usage?– Discuss tomorrow.

• What else?

• Have I missed other sources of funds?

More generally

• Moving on from my “follow the money” games.

• Do we agree what the problem is?• Do we agree that there is a problem?• Can we describe possible business models

for ISPs?• What constraints will regulators impose?

More tomorrow

• Two talks and a discussion about a research agenda.