Yesterday was a day of thirds for me. Two thirds good, one third not so good. In the first two thirds, I attended a roundtable discussion on RSS hosted by MediaFed, a provider of RSS feed tools and services.

It would have been topped off with an excellent three-course meal had I not had to leave for another meeting after the starter (so only one third of a lunch for me, and those that know me well will appreciate how I grieved for the loss of that sticky ginger pudding).

Ahem, but I digress. The purpose of the first discussion was to get some representatives from the UK publishing industry around a table to discuss their current implementation of RSS feeds and how they expect the platform to develop in the future. Before I summarise the points of the discussion, I think it would be useful to summarise what I think are the key RSS requirements from both readers and publishers.

What users want
1. Many people who have embraced RSS feeds are now suffering from feed overload. They want a way of pre- or post-filtering the content so they don’t have to wade through so much stuff.

2. People who have not yet embraced RSS feeds want platforms like widgets that allow them to do or read useful stuff on their desktops or on their social networking pages without having to do anything other than click a download link. In other words, products that already have the RSS feeds built in.

3. They almost certainly don’t want adverts; they definitely don’t want advertising content masquerading as editorial.

4. They may or may not want partial (headline and intro only) or complete feeds; my guess is that most would like the option.

5. They want to get their content from a range of sources of their own choice combined in a single feed.

6. They want feeds they can customise at source, eg by conducting a keyword search of a news archive and being able to subscribe to a feed of the results.

What publishers want

1. To make money from feeds, either by directly advertising within them, or by using them to drive people back to pages that contain advertising.

2. Most publishers probably still prefer to drive people back to their sites by only offering partial feeds.

3. They want detailed metrics on their feeds; who, what, when, why and how.

4. They don’t want their feeds to be used commercially by third parties ie they don’t want anyone else making money off the back of their content.

5. They want to enhance the user experience, but they also want to promote their brand.

Now to the discussion, with the caveat that this is not meant to be a detailed report, more a mix of points made at the meeting combined with my own thoughts.

If people are reading content with their RSS readers via mobile phones, email clients, aggregators etc, what role does the website actually play in the production of content?

The end of the home page? Well, not yet; I have personally seen no evidence of RSS draining away a significant amount of traffic from the home page of Journalism.co.uk and none of the publisher representatives present had either. Which is not to say that it isn’t happening to some degree, or will not increasingly happen in the future.

The Guardian, which currently only offers partial feeds (headline and short intro) is planning to offer the option of full content, with adverts, and hopes that links within that full content will drive people back to the site. The good news, according to MediaFed, is that there is some evidence that people will still click back to your site even if you offer full articles in your feed.

Feed take up is also an issue; The Guardian believes somewhere between five and eight per cent of its general news audience is using RSS. For other publishers that concentrate on technology titles, it’s naturally higher – about 10 to 15 per cent.

I expect that feeds and feed readers will become increasingly more sophisticated and will effectively serve up the equivalent of a web page, complete with channel tabs, multimedia content, advertising, related links etc. But the originating web site will still exist as a central hub to produce and gather content from readers and slice and dice it across various platforms.

What is the future of ‘RSSlets?’

RSSlets are essentially social bookmarking tools that link a post in a feed to services like Digg, Del.icio.us etc, or allow you to email the story link to a friend, or allow you to comment on it. Basically a viral marketing tool, probably still mostly being used by ‘super-users’, ie technologically and web savvy people.

According to MediaFed, readers will use such tools even if you only provide a partial feed. And a lot of publishers fail to include a link to subscribe to their own feed, within their feeds.

RSS Customisation: Are dynamic RSS feeds the way forward in providing the subscriber with the level of detail they seek?

For example, a site visitor conducts a search on a site using a particular keyword and then subscribes to a feed of the results. The Sun offers this facility. Feeds built ‘on the fly’ like this can put a lot of load on a host server if the site is very large with a lot of diverse content and a lot of users, and the other disadvantage is that subscribers may never see other content that might also interest them that they could not have anticipated when they created their own customised feed.

For publishers, this highlights the need to include related links in feeds, especially full ones. It could also provide them with some interesting statistics on what really interests readers and even flag up ideas for applications such as desktop widgets telling you, for example, what the weather will be like on the ski slopes of France (France + ski-ing + weather).

Such a service really puts readers back in control, being able to drill down to such an extent that they might only receive a handful of posts, infrequently, in a particular feed. It is also a service that lends itself to mash-up feeds – aggregations of feeds from numerous sources into a single feed (more on that later).

Full or partial content: what are the benefits and limitations of these formats in driving traffic back to the website?

If your site relies on display advertising revenue, then there is the risk that full content may reduce clickthroughs back to your site where the advertising resides, although MediaFed suggested that clickthrough rates from full and partial content seem to be about equal. To be sure, you might need to sell and place advertising within the feed itself. This might be a hard sell; there are still top advertising agencies out there who have absolutely no idea about RSS. It’s also probably a good strategy to build links back to other content on your site within each article in the feed.

If you present more content in a feed, more often than not it will be read, especially if it contains imagery. If you just provide a headline and standfirst and they are particularly descriptive, then a reader might decide that that is all they need to know about the story, and fail to click through.

SEO (search engine optimisation) perspectives: is there a set of best practice rules to enhance feed distribution and subscriber take-up?

1. Avoid jargon wherever possible – ‘feed’ is better than ‘RSS’.

2. Put subscribe icons everywhere on every page of your site.

3. Use recognisable icons – http://www.feedicons.com/

4. Use widgets and social networking platforms to promote and distribute.

5. Ensure there is a link to your RSS feed in the meta tags of every page.

6. Include subscription links within your feeds and title them clearly so people know where they are coming from.

RSS portals: are feed aggregation services the future for RSS; will we want to pre-filter individual feeds to create a single valid output?

In the absence of customisable feeds, many users would like to be able filter a feed for relevant content, either by keyword, or by category (categories are a bit like tags, commonly used on blogs to group content – if you view feeds with categories attached to posts in IE7, they will display in a side panel). Better still, they would like to do that with a number of feeds from different sources, and then combine them into a single feed.

There is currently a paucity of online services that allow you to create pre-filtered aggregations of feeds – Feed Rinse is one (although when we tested it we couldn’t get it to produce a valid feed), Yahoo Pipes is another.

But there are plenty of aggregation services and if more feeds become customisable at source, publishers can expect their content to be mashed up with those of their rivals.

RSS mash-ups: who is using my data? RSS makes it so much easier for content to be sliced and diced and morphed into other feeds. Are we ‘tolerating’ this at the moment since most of the views still come via the web site? Who owns my data?

Clearly the publisher owns the data and most will have licence terms in place regarding the use of this data. Re-publishing someone else’s entire content without permission and for profit is clearly going to attract the attention of the copyright lawyers.

But, as was the situation with free versus paid content, the readers will drive the agenda here. Those that try to prevent their (partial) content from being mashed up with that of others may well lose out to their competitors. Old news notions about exclusivity and getting the scoop no longer apply online; you can be last with a story and still get the majority of the traffic. Obviously the easier it is is to slice and dice your content in as many ways as possible, the more likely you are to win the numbers game.