Discussions

News: Ten things programmers might want to know about marketers

  1. Seth Godin wrote up "Ten things programmers might want to know about marketers," trying to bridge the gap between the engineer and the marketing dork who spoils his effort. Quick summary: marketing is unrepeatable, and indeterminate. The ten things are:
    1. Marketing is not rational. Programming is. Works the same way every time. Marketing doesn't, almost in a Heisenbergian way. If it worked before, it probably won't work again.
    2. Marketing is even more difficult to schedule than bug fixes. Marketing expenses are easily timed, of course, but the results are not. That's because there's a human at each end of the equation.
    3. Most marketers have no clue whatsoever what to do. So we do unoriginal things, or stall, or make promises we can't keep.
    4. Just because Sergey is both a brilliant programmer and a brilliant marketer doesn't mean that all brilliant programmers are good at marketing.
    5. People often prefer things that are inelegant, arcane or even broken. Except when they don't.
    6. Truly brilliant coding is hard to quantify, demand or predict. Same is true with marketing.
    7. There is no number seven.
    8. Unlike mediocre programmers, mediocre marketers occasionally get lucky. When they do, they end up with a success they can brag about for a generation. But that doesn't mean they know how to do it again.
    9. Just because some marketers are dorks doesn't mean your marketer is a dork. Some programmers aren't so great either. Be patient.
    10. Without marketing, all your great coding is worthless. Push your marketer to be brave and bold and remarkable. Do it every day. Your code is worth it.
    Note that Mr. Godin doesn't address the internal marketing that takes place when programmers are writing applications not for public consumption - but even there, some of this is worth keeping in mind, especially number seven.
  2. We all love to bash on Marketers and marketing. They are the scourge of American society. They're in a dead heat with Lawyers as folks favorite profession to hate. However. As they say, you can hate Lawyers all you want until you need one. Marketers are on the cusp of any business. We all curse their name and love their work. Any time you see most anything on the commercial market today, you can thank a marketer for making it happen. Marketers are a reality today that we can't get away from, no matter how much we'd like to cover our ears and go la la la, while remaining quiet and secure in our back rooms. Rather, it's better to embrace the Marketers and Salemen who affect you directly, educate them to try and not get them to make false promise that end up on your desk to fulfill, and listen to them better understand what little thing may well help them land that next big deal.
  3. Actually, marketing is very easy to measure if you know how and are willing to invest in the tools to measure it. For example, you can measure a branding campaign by conducting a survey before, during, and after the campaign. Lead generation campaigns are easy (assume that all leads from a show are qualified, count # of leads). It's up to the sales team to come up with a model for determining if the ratio of conversion from lead to oppty makes sense. Obviously, the web makes it easy to start putting metrics in place, such as click-throughs, actions, etc. All of this stuff is v. easy to do, but a lot of marketing people don't do it, mostly because they are afraid they will be clubbed for it. Biggest thing is to put in a culture of *doing* the metric so as to get a baseline before you start using the metric as a club. You then incent people for changing the metric, not the metric value itself. e.g. let's say you get 10K hits to the site. You establish targets for % changes to the site each month. Same for things like cost per lead. A lot of marketing people don't do this, but then again, a lot of devs don't want to learn unit tests. Just a matter of training and expectations.
  4. 8. Unlike mediocre programmers, mediocre marketers occasionally get lucky. When they do, they end up with a success they can brag about for a generation. But that doesn't mean they know how to do it again.
    Hrm. I can think of more than a few examples of where mediocre programmers got a good repution for a bad piece of work that was sold right. Both in the public sphere as in the corporate environment.
  5. Dilbert Says it All[ Go to top ]

    Dilbert Sums up the Marketing - Technology relationship perfectly. "Our product is Beige. It uses Electricty". Paul, Technology in Plain English
  6. Marketing is not rational. Programming is. Works the same way every time. Marketing doesn't, almost in a Heisenbergian way. If it worked before, it probably won't work again.
    Where are all these programmers that write software that works the same way every time?
  7. Where are all these programmers that write software that works the same way every time?
    Very nice catch! Send them here, I need one, too.
  8. Marketing is not rational. Programming is. Works the same way every time. Marketing doesn't, almost in a Heisenbergian way. If it worked before, it probably won't work again.


    Where are all these programmers that write software that works the same way every time?
    Precisely. Seth seems to know eveything about marketing, but he doesn't know that: 1) Programmers and programs are two diferent things. Programs work the same way every time they have the same inputs, programmers don't 3) Sometimes programming is like that. Programming is more like new product development than anything else. See "Agile Software Development with Scrum" from Ken Schwaber and Mike Beedle 6) Agreed. 10) Imagine the workd without computers, software, programs
  9. Programs work the same way every time they have the same inputs
    Ever hear of heap and stack corruption? Same inputs different results. Nice try though.
  10. As someone who worked for almost 10 years for a company that, in part, sold software to measure the effectiveness of marketing campaigns and to plan them (and to reduce costs by targetting only the groups that have a high potential response rate), I'd say that marketing is actually a pretty accurate and scientific area. Maybe web-based marketing for the software development industry isn't because the pressures here are innovation (in J2EE the web MVC framework on vogue changes every 3 months or so. That doesn't happen in retail, utilities, industrial supplies, tourism, ...). Otherwise: 5. The bane of my existence working for the services arm of a software company rather than R&D. If the customer is willingto have the application break when 5 people access it at the same time, or there's a file system error because the price is much higher to do it right, then write it wrong. If the customer doesn't need to know that it will break under these conditions and the daily charge out rate is low compared to the next project, then do it wrong, ... 6. Rubbish. There are automatic code checkers and theorem provers can work fine when the code is done right. Theorem provers can't prove things about code that's designed to break them (Gödel and Turing proved that). As for demanding - that's easy. And Predicting is typically easy too: you know the good programmer and you make sure he's not pissed off. 8. Most programmers are mediocre and the chances of getting lucky have FAR more to do with how the project/application is sold and positioned than to do with the code quality in most small scale cases. Unless you're doing something that requires REALLY good code (e.g. google.com) then don't worry about it. Just get it done under budget - that's the whole point of offshore. Michael
  11. Re: Marketing CAN be measured[ Go to top ]

    Unless you're doing something that requires REALLY good code (e.g. google.com) then don't worry about it. Just get it done under budget - that's the whole point of offshore.

    Michael
    It's amazing the number of developers that actually don't understand this point. Specifically, they don't understand that end users of business applications don't really care or understand the underlying technology and certainly never care about what development tools are used. It's one reason why Microsoft does so well with its good-enough-for-most-things development tools that are Window-only, proprietary, etc. PJ Murray, CodeFutures Software Data Access Objects and Service Data Objects

  12. Rationality is over-rated[ Go to top ]

    #1 is true, but I would say "Life is not rational", "People's desires are not rational" and "Technology marketing is about connecting people's desires with the innovations that fulfill those desires." As for predictability, it doesn't make markets. And considering that all the best developers I have worked with bring a lot of passion, craftsmanship, and inspiration to their work, an argument could be made that rationality is overrated in that field also. ;-)