Programming considered harmful? (Yahoo)

If I were running Yahoo I would say “We have all of these users, most web services are pretty bad and have glaring missing features, so let’s build audience by creating more and better services.” This analysis by a hedge fund, however, shows that Yahoo was, as a percentage of revenue, a bigger spender on R&D and product development than either Google or Apple (see slide 14).

How is this possible? If you have a huge audience and competent programmers and a world full of unmet needs, shouldn’t one be able to make money through coding?

Slide 40 is also confusing to me. Instagram and tumblr were acquired for about the same price and have the same number of users. One is worth $35 billion, according to the slide, and one is worth $0. Why the difference? Presumably Instagram has a lot more revenue per user, but why?

Could the answer be that hiring programmers in Silicon Valley is not cost-effective anymore for ad-supported businesses? (Google, presumably, being an exceptional case due to its market power.) I talked to the CEO of a 400-person company involved in online publishing yesterday. He said that he had shut down the company’s California office. Web development is now being done out of Vietnam where a programmer whom he considers to be high quality costs $15,000 per year. For mobile development… Barcelona.

Finally, look at the stock compensation graph on slide 12. While Yahoo achieves nothing, except for continuing to hold onto Alibaba stock, investors pay the employees $420 million per year in dilution via stock grants. The CEO is taking $365 million from the investors (page 54; Sheryl Sandberg would no doubt point out that the CEO would be paid a lot more if she were a man), whose board members were dumb enough to tie only 3.3% of total comp to the company’s performance (see my economic recovery plan for the U.S. on why governance of public companies is so bad). The CEO is described as incompetent but slide 56 shows that personal finances are being managed brilliantly (i.e., the CEO has been selling Yahoo stock as fast as possible). The classic paper “Unskilled and Unaware of It: How Difficulties in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments” suggests that an incompetent CEO would be unlikely to sell based on his or her own incompetence.

Readers: What do you think? How is it possible that with such a large audience, which adds tremendous leverage to even the simplest coding achievement, Yahoo is not successful?


  1. Bill

    December 18, 2015 @ 1:28 pm


    Yahoo never figured out what businesses they were and are in and exploited it. They were and are in the “newspaper business” mostly and this is just not a big money business. Plus they never added to this business model and now many of the other opportunities and offshoots of that are all taken by others. So they lose. Yes they were early but who cares.

    But what is even worse is they have not exploited their strengths like Yahoo Finance. There is huge money in investing and helping Joe Public invest and manage his money but they are ignoring him. Oh well.

  2. paddy

    December 18, 2015 @ 2:33 pm


    Programmers can program their way around everything except “general cluelessness”.

  3. jack crossfire

    December 18, 2015 @ 5:40 pm


    Every conceivable product now requires software, so any employee hired for anything is now considered a programmer when they’re really orange juice makers, bankers, auto mechanics, & farmers. Investors want to see every industry requiring software move to San Francisco & every employee called a programmer.

  4. ianf

    December 18, 2015 @ 6:09 pm


    Not much of a prophet, least of all in matters of such financial scope, I can nevertheless see a merger of Yahoo! and AMZN shimmering in the near future. Once joined, YahooMail becomes A KINDLER, Gentler Mail and is put on all Kindles etc as the default (if not the only) option: a free, ETERNAL[*], unlimited storage size mailer and transparent cloud archive.

    This would permit AMZN to “peak into” our mail-borne interests (aka keywords) just as Gmail does today, and then serve buying hints right within the client (presumably in a non-intrusive, optional fashion).

    As an aside, email is long overdue a paradigm-shift level overhaul for the limitations, requirements but also ergonomic advantages of mobiles and tablets. The Unix single function app model of yore has outlived itself. Because, frankly, in the age of always present pocket/portable devices that sync themselves to “clouds,” I don’t see why one should keep composing text online in different target sites’ own, often quirky and irrevocable text submission dialogs, when the same could just as well be written asynchronously in mail, and simply posted to a specific target from within it (early Eudora and other mail clients were able to NNTP-post to named Usenet newsgroups as well, so what’d prevent posting via HTTPS over today’s SMTP? Receipts and confirmations would be added to the local post, no need for separate, time consuming “Comment published!” messages.)

    The same goes for reading selected target sites… some mail clients already are capable of rendering RSS feeds as were they serial messages in a mail file.

    Thus all that one created/ sketched/ posted/ whatever would automagically become part of a cumulative sequential private offline/cloud “extended email, etcauthoring record (goodbye lifestreaming web services that collate one’s blog posts and comments from different places provided each source site is first laboriously identified to the service). A READY-INDEXED, on-demand tagged, keyword-cloud-summarized, post-edited to suit writing record that’s up-to-date and available for advanced/ RegExp searches from the moment it’s been modified and refreshed. An email client that allows adding hypertextual links between arbitrary elements within it and to the web at large; mail front-end that works just as well as a note-taker, a to-do list maker, outline composer, “commonplace book quotes” recorder, etc. After all, what are text notes but mail messages sent to oneself?

    Thus the combo extended-mail-tablet would become more of a personal “Dynabook” device, with one major GUI for text/ picture/ etc input, and adaptable reading/ viewing UX. Not solely for ebooks, but for web etc presented as were it (e.g.) an ebook.

    I’d have preferred this to be on an  iDevice, but if that Amazing Yahoo! Kindle gets there first, more power to them.

    [^*] Gmail is limited to 15GB + 9 months life past the last account access date.

  5. G Close

    December 18, 2015 @ 6:52 pm


    Yahoo fumbled the ball when they outsourced their main profitable business (search), to Microsoft. This was during the Carol Bartz era, eons ago.

    Online news and entertainment with ads to support it are a very unprofitable business, compared to ads for search. And even the latter may not be very profitable for more than 1-2 players with very large market shares, due to the investment required.

    Marissa Mayer never had a chance, although her negotiation ability for compensation is clearly first-rate. I’d say this is fairly common knowledge in the Bay Area. Should’ve taken the Microsoft bid and ran, but hindsight is 20:20.

  6. michiel

    December 19, 2015 @ 3:51 am


    With respect to the difference between Instagram and Tumblr; Instagram is now strongly tied to Facebook’s infrastructure. Since it’s almost solely used on mobile phones it provides Facebook with a ton of personalised behavioral data. It caters to people’s vanity, encouraging them to display a glossy, idealised version of their lives; this is a very ‘natural’ environment for advertisers.

    Tumblr is basically the picture-centric rebirth of the old Usenet alt.* hierarchy. It’s anonymous, fragmented, and its users stubbornly refuse to present themselves as an easy target audience.

    I don’t know any hedge fund managers, but I picture them as the kind of people who use Instagram a lot; they probably don’t know anyone who uses Tumblr.

  7. Brian

    December 19, 2015 @ 7:11 am


    &ianf awesome idea which someone would make it.

  8. Calico

    December 19, 2015 @ 3:31 pm


    Innovation and out-competing competitors in SV is extremely IQ based.

    As computer programming is effectively a technical skill, only really pay attention to the M part.

    Once you apply this same graph to java/c++/javascript programming capabilities, along with effective knowledge of implementation of data structures and clever dynamic programming you basically find that R&D innovation is nigh impossible for companies that can’t find the right talent.

    There’s a difference between R&D and typical somewhat complex web development. With the latter, a smart(yet not genius), hardworking person with a good knowledge of the latest tools (and perhaps graphic design tastes) can be hired anywhere in the world that has a functional educational system alongside talent locators.
    With mathy R&D , there’s so few people capable of really doing anything truly innovational that if a company doesn’t find the right talent, its mostly there to say one innovates.

  9. Calico

    December 19, 2015 @ 4:22 pm


    That may be less of a critique of yahoo, and more of a generalist “This is the first thing someone needs to look at”

Log in