ananddass

Disclosure: I am on the B.D & partnerships team on Postmates.

We haven't shifted focus away from restaurants. We have about 3000 Plus partners today of which a majority are restaurants. We've always done more than restaurants. Food is most of what we do but there are also grocery, apparel, electronics, essentials retailers on Postmates.

It really depends on the order distribution of distances for a specific store but on average its ~7 minutes from pickup to dropoff. Regarding difference in temperature, the coffee travels in insulated bags and so we aren't seeing significant differences. We are monitoring customer satisfaction scores on these orders and will tweak as we go along as we find opportunities.

Actually, Postmates & Starbucks have been working on this partnership for a while now. We began operationally testing this in the last couple of months. As with all tests, there were kinks that needed to be worked out and new problems to be solved. Transporting coffee, spill-proof, at scale was one of those.

We were optimizing for speed so didn't really shoot for optimizing costs as much. It cost us about $6.61 in material costs and $5.25 in laser cutting time.

Microsoft needs systemic changes. There may some truth in the fact that ballmer is to blame. But a new ceo will face the same challenges. Markets are irrational.

You may want to check out what Jobs had to say about this. See quote from 7:12PDT : http://news.cnet.com/8301-13860_3-20006442-56.html

The goal was to understand, if we are there yet with respect to the readiness of iPad software. Bucketing iPads as a consumption and leisure device is short selling its potential.

It is true that app quality has a lot to do with the experiment. I was surprised by how many of the problems could be solved if the apps upped their game. Hardware limitations were secondary only to the software experience.

That would be a good analysis to do as a next step. The new Microsoft ads make it a point of highlighting that differentiation.

aaah Ok. Its easy to buy something but hard the switching effort and learning effort is higher so retention is lower. Interesting.

So one of the things interviewees mentioned is that there was a difference between "micro moments" and "full blown work".

Writing a paper is full blown work that they know they cant do on a mobile device. But they expected a lot of the micro-moments to work.

Yikes. Fixed the fonts to make it more readable.

Curious as to why would engagement and retention be low? As regard's pay per download models...looks like "Attach" business model's are becoming more popular.

"Attach" means charging for something outside of the app store. Dropbox, Uber, Square all do it. I remember seeing a Chris Dixon post about this on HN a few months ago.

Sorry, I should have added more references in that connected the dots better. A lot of capital and talent is flowing into solving the problem. Thus, we are likely to see some big shifts. --$4 billion invested into "business/office" software on iOS since 2010 http://readwrite.com/2013/06/25/a-new-golden-age-of-producti... --Mailbox's exit to Dropbox --Quip openly calling for a change to the word processor https://quip.com/blog/introducing-quip

Yes. The keyboard problem is a big one. It was so painful that I had to buy a bluetooth keyboard just to work with my iPad. That's a hard one to solve until apple comes up with something better user input device. (e.g. thalmic labs, leap motion etc).

After I bought my keyboard, the next problem I hit was the inter-app communication issue you raised. That's a software problem and solvable.

That is precisely the issue-where are our resources best spent?

Fair point Chris. User testing has been changed to usability testing throughout the article. Referenced your comment on HN

Unfortunately, I couldn’t change the title as it would break the url which has been reblogged and distributed.

It makes sense to pass files by reference rather than by value. This integration example would allow you to send files as urls via the SMS protocol instead of sending blobs of data.

Mobile network bandwidth is certainly a constraint in the US (cellular networks throttling high data usage subs is a proof of that). Viewed in that context pass by reference makes greater sense.

Jordan and team are good friends of ours (Filepicker.io team). We are trying to solve similar problems but approaching it from different directions. Its going to be an interesting future. :-)

Any thoughts on the boundaries of A/B testing pricing? It can be construed as very sketchy if you show one set of visitors a certain price and some other another price point.