2 posts tagged “turbotax”
I have faced this question a lot of times when designing software for product designers? It is very easy to say "let us automate this and that?". Given the state of software development and the number of bugs every software has (just read the report on Boston Globe about US Airways flights being delayed the last couple of days because of glitches in the new upgrade of check-in software), you wonder how far you can take "automation" before it gives a negative ROI?
Don't get me wrong - I love automation. I am a mechanical engineer - I have some idea how my automatic transmission works - but when I am driving my car that is something I just want to work. What about doing charts or calculating trend lines or the myriad of functions in Excel or doing my tax return using Turbotax - I love the fact that this is automated and it just works.
But there is a place for automation as well. It better work all the time, otherwise there is nothing that kills the productivity more because the user will have to first undo what automation "did" and then figure out how to do what he wanted to do in the first place.
Some cases of automation gone wrong is MS Office
1) "paper clip man" - because he tries to get in my way by offering to automate things when I don't need it.
2) When Word offers me to merge reviews together. Given that I have been bitten in the past, I always say No. Because when I am getting multiple reviews from others and if it does anything wrong (and it has), I will have a helluva time undoing what it did and then doing it the way I wanted to do it in the first place.
In CAD (which is digital clay because what you can create is limited only by the user's imagination) I have always felt that automation should always keep the user involved in what it is doing, so that if the computer did something wrong, the user could always correct it and still get the desired results. The latter becomes a lot more harder if the user had got accustomed to the automated way of doing things without having a clue of how automation did what it did. But at the same time, we always talk about removing the CAD overhead for our users and letting them concentrate on their product design. This can be done by making CAD more easier to use and this would require more automation of tedious and "Cadeze" tasks, but we still have lot of work to do here.
I read an interesting blog post by Kathy Sierra on Creating Passionate Users titled "Are our tools making us dumber?". Very interesting read indeed and the 77 comments that it has attracted is worth reading as well to get different perspectives.
Well, I have been silent for a few days primarily because I am busy attending the Voice of the Customer conference here in San Diego. The weather is nice in the upper 60s, it is sunny during the day. But the conference is really good that I have not had the chance to go outside other than for dinners when it is dark.
There have been some great presentations and some bad ones as well. There was a very good presentation by Heather Crothers from Intuit on how they immerse themselves in the life of the customer. Intuit seems to get it and no wonder they are so successful in selling a product that we all have used - TurboTax.
But one thing I learnt from the conference is the concept of shared interpretation vs. sharing data. We live in a world where most of the communication happens over email. We share data that we have gathered from customer visits, experiments that we have done with others in the organization and our conclusions. All this is good, but not sufficient. If you want to make a change based on what you have learnt, then you need your organization to have a shared interpretation. Sharing data alone will not cut it. It will be interpreted in multiple ways. You need to make sure everyone is interpreting the data in the same way - hence the name, Shared Interpretation. Sounds common sense, but hey who was it that had said "Common sense is not that common"
