Categories
Uncategorized

Internet-based software and the art of the pleasant surprise

I had to deal with some personal e-mail today, and I noticed something…significant. As much as I’ve advocated for app-like websites, and I do believe they are the future, it’s exciting to see a tipping point. But before we get there, let’s take a couple of cases of app versus app website, and walk through some analysis.

  1. Desktop/laptop email. Traditional email, sadly, isn’t keeping up with web offerings, which is unfortunate, because it is a great opportunity for us, as users, to have a ‘single inbox’ view that’s useful for choosing and prioritizing our work. But since Google, Microsoft, et al. are moving to appointment systems that are not interoperable, this makes desktop mail have difficulty keeping up, unless it ties into the website view.
  2. Phone and tablet email. Sadly, the same story; and to make matters worse, the main platform vendors have their own apps, which don’t handle the other platforms.
  3. Small apps that reflect websites that have moved onwards. I can’t disclose too much, as one of them is a (potential) client, but sometimes apps stop working, and are no longer economically viable. Purely launching to the website view can absolutely be a useful escape hatch for apps like this; it maintains the presence of the app in the app store, and reduces duplicating functionality in the app.
  4. Some app store platform owners, e.g. Apple, demand revenue for all in-app purchases. This might be an escape hatch for that as well.

So, here’s the pleasant reminder and a concrete example of what I’ve been pointing out at work for years: Sites can be updated to surprise and delight the customer *without* installing or even developing new app software, and if done carefully, the new features can be self-introducing. This is also a way to escape huge development and release cycles, as well as onerous pre-launch training, if you don’t break anything else!

This would be absolutely outstanding for healthcare and government, which are currently far too conservative about introducing software change.

Today’s beautiful example of this is Yahoo e-mail. Yahoo recently introduced a “subscriptions” view of the inbox…and it’s really very good.

First, the feature is introduced by little yellow ‘new’ tag that points to the feature. This is *excellent*, and uses the ‘show don’t tell’ approach from story telling, film making, and good signage from roads. Sadly, I was too excited to see it and clicked it immediately, so I don’t have a screenshot.

Second, the feature itself shows newsletters by subscriber, frequency of posting, and has the unsubscribe feature prominently featured. This is a great example of user experience design (UX), specifically use-case focused user interface (UI) design looks like.

Notice the grouping of views; the subscriptions view is selected, and the items in the view are use-case specific. Click to view (the primary affordance), and a button for click to unsubscribe.
Notice the grouping of views; the subscriptions view is selected, and the items in the view are use-case specific. Click to view (the primary affordance), and a button for click to unsubscribe.
And here's what happens when you hit unsubscribe. Not only does it do the "Are you sure?" style dialog box, but it also tells you what it is going to do, and what to expect. This is a great example of what to do when you can't pull off the "show, don't tell" approach.
And here’s what happens when you hit unsubscribe. Not only does it do the “Are you sure?” style dialog box, but it also tells you what it is going to do, and what to expect. This is a great example of what to do when you can’t pull off the “show, don’t tell” approach.

Congratulations to Yahoo! Most of my previous colleagues and acquaintances from the mail team have moved on, so I don’t know who to extend those congratulations to personally, but this is fine work.

Categories
Uncategorized

Large Review Finds Good Evidence for Masks, Distancing in Stopping COVID-19 | MedPage Today

Good news: masks are a huge help.

A side “how to read this” note: calls for randomized trials is really people falling back on their comfort zone of “that’s how we test medicine” but it is *not* necessary or even possible for many interventions, so requiring randomized double-controlled studies introduces a different kind of bias: it excludes interventions where that isn’t possible, and adds a bias towards medications instead of prevention. This bias drives the costs of healthcare up, in that treatment rather than prevention is almost always more expensive. Other controls are possible, and observing after the fact “natural experiments” and meta-analyses like this also provides huge amounts of information that can be acted upon. In a pandemic, especially, wait and see on known effective preventative measures kills people. When the intervention is low risk and high payoff, just do it, and keep taking data to confirm or disprove whether it works!

That said: read the attached article.

www.medpagetoday.com/infectiousdisease/covid19/86812