Comment by gravitycop | Hacker News Reader

Comment by gravitycop | Hacker News Reader

Parent Full thread

nowadays if you design anything at all that has an end-user component you'll spend at least as much time at making it jerk-proof and doing all kinds of abuse analysis

http://www.google.com/search?q=%22always+blames+his+tools%22

Echoes of Kozmo: http://money.cnn.com/2005/09/14/technology/kozmo_redux

No more 50-cent deliveries [...]

Customers of Kozmo.com who fondly remember getting single pints of ice cream delivered on a whim won't get off that easy this time. MaxDelivery has a $10 minimum order, and there's a $4.95 delivery charge for orders under $50. Siragusa said this is to keep customers from abusing the system, as they did with Kozmo, and he says MaxDelivery's average order size is much different than Kozmo's.

"Someone ordered a pack of Mentos. Then two hours later they'd order a pint of ice cream," said Siragusa. "People did this because there was no penalty. Kozmo had a number of orders that were unprofitable."

Kozmo's founders designed its system to reward customers for making frequent low-value orders. When the customer base responded in a way that should have been predicted by the Kozmo founders, instead of recognizing and fixing their error, those founders instead called it "abuse".

I'm perplexed by your implied comment that having to spend time making your product jerk-proof and performing abuse analysis is only done by poor workmen.

It seems to me that work must be done to create value, and work must be done to make the system idiot proof (despite the continuing ingenuity of idiots), and this is true regardless of the talent and abilities of those producing the service or product.

I'm perplexed

Work (value creation) can be more or less efficient. As stated, an inefficient workman blames his tools, instead of the way he is going about his work. In this case, and often in other cases, the customer base forms part of the toolset. The workman is saying that his tools are not doing what they are supposed to be doing, and that it is their fault.

Here is more on this phenomenon: http://en.wikipedia.org/wiki/Locus_of_control

Internals tend to attribute outcomes of events to their own control. Externals attribute outcomes of events to external circumstances.

In the present case, the customer base would be seen as uncontrollable (only susceptible to external control) by the inefficient workman, and controllable (susceptible to internal control) by the efficient workman.

work must be done to create value [...] regardless of [...] talent and abilities

Yes. But do you mean, as Marxists contend, that variance in ability plays little or no role in variance in the efficiency of the work?: http://en.wikipedia.org/wiki/Labor_theory_of_value#Marx.27s_...

I can't figure out if you're just a very clever troll or someone that has something to contribute, if the former, then congratulations, well played but it does not really help the conversation, if the latter then maybe you should try to make your point in smaller steps so that lesser minds like mine can make sense of what you are trying to communicate.

In this case ... the customer base forms part of the toolset.

I do not subscribe to this theory. Toolsets can be changed, improved or mastered. I can do none of these with my customers. I can only do it with the things I use to create my services. I use tools to create services to assist my customers. Calling my customers "tools" does not help me.

The workman is saying that his tools are not doing what they are supposed to be doing, and that it is their fault.

I disagree with this analysis. The workman is saying that there are (usually a small number of) customers who require that the services be more complex in their implementation than might otherwise be expected. This is not the same thing. I know that around 80% to 90% of my customers never provoke the inbuilt protection systems in my code. The protection systems account for about 50% of my code. I have measured these things because I work in the defense industry. I know that around half my code is there because up to 20% of my customers don't use the system the way it was intended to be used. Whether this is lack of training, lack of awarenes, or malice doesn't matter. I don't "blame" them, but I know that if my customers only did what they were shown in training, and didn't do what they are trained not to do, 50% of my code would be unnecessary.

But do you mean, as Marxists contend, that variance in ability plays little or no role in variance in the efficiency of the work?

Of course not. I mean that for every worker, some of their work is producing features and facilities, and some of their work is protecting against inappropriate actions. The balance between these two might change with the abilities of the workers, but I suspect not much. It's a pretty constant ratio across my programmers. To assume what you seem to have assumed from what I have said is, again, perplexing. Perhaps we constantly talk past each other, but I'm finding it difficult in general to understand your mindset.

>> the customer base forms part of the toolset.

> Toolsets can be changed, improved or mastered. I [cannot] do [this] with my customers.

You cannot provide any value to the world that would involve a different customer base? After Kozmo re-emerged as MaxDelivery, it changed its customer base (one of its tools in value creation) by starting out with only high-density neighborhoods of Manhattan (and it continues to change its customer base as it gradually expands its delivery areas). It has mastered its customer base (again, one of its tools in value creation) by learning how that tool reacts to various delivery deals, and by employing improved delivery deals. Please see the article linked above.

For at least many decades, bike-sharing advocates and politicians have been failing to efficiently employ the tools at their disposal, one of which is their customer base. That tool, the customer base, has been blamed over and over again as "the problem" as bike-sharing advocates and politicians have repeatedly employed essentially the same defectively-designed policies.

I don't think that you can compare a broken business model with abuse of a system by spammers and others hell bent on destroying some of the nicer sites that well meaning teams have developed.

By that analogy usenet was 'asking for it' because it was open, worked and actually contained great content. Of course that meant that it was ideal for the jerks of this world to try to peddle their wares, eventually resulting in the near destruction of usenet.

On the web the relationship between the number of people that you have to employ to look after abusers vs the number of people that are actually productive is a good way to measure your success....

the relationship between the number of people that you have to employ to look after abusers vs the number of people that are actually productive is a good way to measure your success.

Yes. That is what I said. The workman with internal-locus-of-control improves his value-creation system based on feedback, and is seen as more successful. The workman with external-locus-of-control blames parts of his value-creation system for acting (in his eyes) wrongly, and is seen as less successful.

Recognising that customers behave a certain way and taking that into account is valuable. Blaming the customer for behaving that way is of little use. You have conflated people recognising that people behave in ways that cause difficulty with blaming them for doing so. Not helpful.

Separation of concerns is as useful in customer identification as it is in writing software.

If I change my model to change my customers, I am changing my customers, not changing my toolset. You are deliberately confusing/conflating two fundamentall different concepts,and I believe that while you personally, and perhaps others, may derive some benefit from it, doing so inhibits effective communication.

http://xkcd.com/169/

I agree that blaming the customer base for the failure of the bike-sharing schemes is wrong, and I agree that it may be that changing the target customer base is the only effective solution, but do not agree that calling that "changing your toolset" is helpful.

Reading again about "Locus of control" and putting that into the context that you have further provided, I think this is still a question of attitude, and your comments remain unhelpful at best, and at worst are actively misleading.

Viewing the customer as a part of my toolset does not make them internal. A defining feature of my toolset is that it is under my control. The customers I have are not under my control.

Viewing the way I generate my customer base from the collection of potential customers is largely under my control, but that is not the same thing. My customer base is more under my control, but is still not part of my toolset.

I have control over what I produce, and viewing what I produce as a mechanism for selecting customers is useful. Realising, and explicitly using, the fact that what I produce will implicitly determine the customers I get is valuable.

This is not achieved by viewing customers as part of the toolset.

Reply on news.ycombinator.com