Humility and taramasalata have much in common. Neither is much of a meal in itself.

The first is but one ingredient in a pie, the other more normally consumed with pitta bread. And both are seldom more than appetisers to the main course.

So I will rapidly ingest the option encased in pastry, hopefully without suffering too badly from indigestion.

Because last week I argued apps were not the answer to extracting payment from people wanting to read news on their mobiles. But now I think they just might be.

I also stated newspapers would never be able to rely solely on advertisers to generate sufficient revenue to profitably distribute their content online. Again I suspect I might be wrong, at least where mobiles are concerned.

I had forgotten my Marshall McLuhan. The medium is indeed the message.

And your computer is not your mobile.

The fact is the significant difference in screen sizes imposes separate constraints upon how content can be both presented and read. But that difference also creates an opportunity.

The internet, for reasons touched on elsewhere, is largely ineffectual at conveying display advertising. Unlike print, people experience tunnel vision when reading on screen. And even though banners are placed above, next to or even within editorial content, they are mostly ignored.

Consequently the price a newspaper can charge advertisers for each thousand readers on its website remains but a small fraction of what print still commands.

However, suppose we stop thinking of mobile as simply being an alternative internet platform, with advertising presented alongside editorial. The reduced screen real estate effectively makes that impractical anyway.

Suppose instead we think of it like radio and television.

With broadcast media, the commercials are intrusive. They interrupt the programming. The only way to avoid them is to leave the room, change channels, or hit the off switch.

So suppose, when you click on a link to read a news story on your mobile, the publisher first serves you a 10 second unavoidable “commercial message”.

In return, there would be no subsequent advertisements interrupting the story itself. And the bytes deployed in serving you that 10 second commercial would not affect your monthly data allowance.

Of course you could opt out of having to receive those advertisements but, as with Spotify, that would require you to pay a monthly subscription.

In other words, provided publishers could charge advertisers a sufficiently high price for each thousand ‘commercials’ served, or sell enough subscriptions to those wishing to avoid the advertisements, or both, the concept could work.

Time for some simple back of the fag packet arithmetic.

As of the first quarter of this year Ofcom estimated there were 8.1 million individuals in the United Kingdom accessing the internet on their mobile phones. Around the same time comScore claimed 48% of all UK smartphone users were accessing news or information via their browsers.

Of course, not all of those accessing the internet on their mobiles will have smartphones, and not all of those wanting news or information will necessarily access newspaper websites.

However, for the sake of argument, let’s say the number doing so then was 3.0 million and, with smartphone sales made since, has now reached 3.5 million. And all forecasts suggest that figure that will continue to grow for the foreseeable future.

Should each of those 3.5 million users visit at least one newspaper website on 15 separate days each month, viewing on average 4.2 pages per visit, then a monthly total of 220.5 million pages would be viewed.

Were each of those 220.5 million pages to be preceded by a ‘commercial’, and were advertisers prepared to pay the £12 per thousand impressions that Jupiter Research estimated was being paid in the UK earlier this year, £31.8 million would be generated each year.

That is a useful sum. But, when divided amongst all national newspaper publishers, it does not amount to much.

Fortunately I think that underestimates the actual revenue potential by perhaps as much as a factor of four. And key to achieving that higher cost per thousand is the ability to guarantee to advertisers that they will only reach their specified target audience.

For example, should I simply wish to target women and were I to buy a colour page in The Sun in order to do so, I would reach 13.8% of all adult females in the UK and, at ratecard, I would be paying around £50K for the privilege.

However, it’s not just women who read The Sun. In fact 54.9% of all readers are male. So you could say that of the £50K I was spending, more than half was being wasted, as I still had to pay for those men, even though they might never buy my product.

More pertinently you could argue I was happy to spend £50K to reach 13.8% of all adult females, and to pay a cost per thousand of £14.60 to do so.

But that still leaves all those unwanted men. Were The Sun able to sell them separately to an advertiser only wanting to reach men and also prepared to pay a cost per thousand of £14.60 to do so, then instead of £50K the page would generate £112K.

In truth few advertisers simply wish to reach “all adults”, “all men” or “all women”. There are usually age, income and other filters they want to apply. So the wastage is normally even greater and the cost per thousand paid if badly negotiated even higher.

Unfortunately for The Sun a single page can only be sold the once, so it cannot profit from offering any readers one advertiser does not want to others who do.

Online, of course, such constraints do not exist. Provided the publisher, at the very least, knows whether the person being served the ad is male or female.

Indeed, the more that is known about that individual, the less the wastage and the greater the potential revenue generated across the inventory. So not surprisingly there is considerable enthusiasm amongst both publishers and advertisers for techniques like behavioural targeting.

Unfortunately, knowing all the websites an individual has visited and what they did when there will still not provide all the answers. The possibility also exists that exploiting such data may soon run foul of the regulators.

Ofcom has just announced an enquiry entitled the Advertising and Pricing market study. “In particular,” say Ofcom, “we may look at behavioural advertising where information on a consumer's online activity is used to target the internet advertising they see.”

Yet the fact remains, to unlock value publishers need to change the question being asked by advertisers from “how many” to “who”, while knowing that they can provide accurate answers.

And that, very neatly, brings us back to apps.

Because, suppose before you can access any newspaper website on your mobile, you first have to download an app to do so. And suppose, in order to download that app, you have to provide some basic information about yourself.

That information might include your age, sex, postcode, occupation and marital status.

You would also be informed that you will either have to agree to those 10 second commercials preceding any content you might wish to read, or else you will need to pay a monthly or annual subscription.

Should you opt to accept advertising you might also be asked to select what categories of advertisement you would prefer to receive. For example, were you to be thinking about buying a new car, automotive might be amongst your choices.

You will, of course, be able to change your preferences at any time.

Given the choice of either paying a subscription or having to provide personal information and be exposed to advertising, many might opt to pay the subscription charge, particularly were they in the habit of reading a lot of pages and the cost to be set as low as £5 per month or £50 per year.

Conversely those opting for advertising could well be prepared to state their category preferences. Using only the personal information provided publishers will be able to target advertising with considerably greater accuracy. But with up to date category preferences those commercials could also be of greater current relevance. And most people will acknowledge there has to be at least some small benefit in trying to ensure the advertisements served might prove vaguely interesting.

However three facts are for certain. Nobody is going to download a dozen different apps for a dozen different newspapers. Nobody is going to pay a dozen separate subs. And nobody is going to take out a sub or agree to accept ads from one newspaper if they do no have to do so for the others.

As far as the first two of these are concerned, one app and one sub is all anybody should need to access every national newspaper.

As for the third, a trade body like the Newspaper Marketing Agency should pick up the telephone to the good folks at, say, Google.

A single highly functional app needs to be developed that can pull in and, where necessary, reformat newspaper content to ensure readability on all mobile platforms.

A central database needs to be created of all who have downloaded the app, and subscription income needs to be allocated to publishers in proportion to both pages viewed and time spent reading.

Advertising would continue to be sold by individual publishers but served by Google. Any unsold inventory could be made available to Google to sell.

In return for their services Google would be paid a monthly fee dependent upon the number of active apps, along with commission for any advertisements sold.

All newspapers would go live with the app on the same day, and all publishers would ensure that their web pages would only download to mobiles on which the app had been installed. A simple check on the screen size of any device making a page request should suffice.

No doubt there is a reason, other than the BBC, why such a plan will not work. But if it could be made to succeed it might then be extended to devices other than mobiles.

And were that to be feasible, with 19.6 million UK residents visiting a Newspaper Marketing Agency member website in June alone, the possibilities become very interesting indeed.