Every company builds native apps for mobile, and web apps for the PC. Mobile web apps are a neglected third option, and are sometimes even being shut down.
But I think this view is mistaken. Mobile web apps are critical. They serve a whole set of use cases that native apps don’t, or serve poorly. Each of these use cases may individually be marginal, but they add up to have critical mass. Developers would be foolish to ignore them.
To begin with, a user may not know that a site has an app. Or they not want to deal with the hassle of installing it unless they’re going to use it regularly. Or they may be okay with installing it, just not now, when they’re in a hurry. Or they may not be okay with the permissions the app wants, on Android.
Some native apps are crippled, missing critical functionality. For example, Simplenote, a notes app, offers revision history, but not on their Android app. If your hand touches the screen and you accidentally delete part of a note, which happens to me all the time, you can’t restore the note using the Android app. This forces me to use a web app on my Android phone.
Some native apps have a UI inconsistent with the desktop version, for no apparent benefit. Articles, an iOS Wikipedia app, comes to mind. It formats and presents Wikipedia articles differently enough from the official Wikipedia sites (both the desktop and mobile versions) that I never feel at home using Articles. I can understand doing things differently when there’s a concrete, identifiable advantage, but needless inconsistency produces confusion and a lack of feeling of comfort with the app.
Developers of native apps also need to remember that they are competing with the browser, and the functionality that comes free with it. For example, I dislike many Wikipedia apps because they don’t offer tabbed browsing. I use this all the time on Wikipedia to follow interesting links in the middle of articles without interrupting my reading, or losing my place in the article I’m reading. At other times, I found myself missing Find in Page, or emailing a link to myself to continue my session on a PC. Native apps miss functionality that comes for free with the browser. This makes me switch to the web app every once in a while.
Some native apps can also have paywalls that are less open than their web counterparts. For example, links to The Economist articles open in a browser, but not in the native app, which demand that I buy the issue .
Users with low-end phones may not have the storage space needed to download many apps. In fact, many of them have space for just 2-3 apps, which means that they can install only critical apps. For them, this whole debate of native vs web apps is largely academic, since they don’t have the option of installing most apps for services they use.
Put together all the reasons above, and the conclusion is that mobile web apps are critical. Each reason may individually be marginal or debatable, but put together, they have solid mass.
Anyone who neglects building a mobile web app, or builds a shabby one, is putting their own convenience ahead of their users’ convenience. And that’s obviously the wrong mindset for a developer to have.
The counterargument is that it’s more nuanced  than that. By sacrificing a mobile web app, the argument goes, companies can focus more on building a better native app. But by that logic, it would be simpler if developers target only Chrome for their desktop web apps, telling Firefox, IE and Safari users to go away. You don’t find anyone seriously pursuing that strategy, do you? Why should the mobile web be any different? Developers need to meet users where they are, rather than demanding that users bend over backwards to make developers’ lives easier.
 Which is a good example of the magazine publishers’ outmoded thinking. It makes no sense to ask someone to buy a whole issue of a magazine just so that they can read the one article whose link they clicked on.
 I suppose the way to settle that argument is with data. What’s the number of page views on the native and mobile web apps? What’s the revenue generated by those users? And if you shut down your mobile web app team, and put those people to work on the native app, how much additional revenue can you generate? Is this more or less than the revenue lost due to shutting down the mobile web app?
Page views, in particular, merit caution — we should also consider second order effects. A mobile web app may have low traffic, but may be the trigger for the purchase.
For example, when I do a Google search for a particular product, on my phone, and click a Flipkart link, and I get a page asking me to download the native app. If I’m using an Android phone, I refuse to accept Flipkart’s permissions to install the app, so I press the back button and instead go to Amazon. If Flipkart had a mobile web app, I might read it, and later, when I have more time, open the Flipkart desktop or native app, check more information, check other products, read reviews, and then buy the product. There might be 1 page view on the mobile web app, and 10 on the desktop or the native apps, but it would be wrong to look at that data and conclude that the mobile web app is not important. That single page view is the start of a process that results in 10 page views and a purchase. Turn the user away at the first step, and you’ve lost the purchase. And make future purchases like likely, because they’ll think twice about using Flipkart the next time.
So, yes, in theory, data should settle the argument, but it’s hard to get the data, and it can lead you to wrong conclusions.