Grapeshot replaces "hard-wired" editorial segments with the flexibility for ad sales teams to create virtual channels which mirror their commercial opportunities and objectives.

Editors create website segments to help users navigate around a site. Ad sales teams often inherit these same editorial sections and try to sell them at premium prices. But it is sometimes hard to match advertisers needs to these rigid editorial channels. Grapeshot introduces a more flexible method.

Humans have limits and it is the journalist's job to write the story, not organise it like a central librarian into categories and taxonomies. So the obvious stories go into special segments like "Technology" or "Lifestyle", but the bulk may still go in "News".

Traditionally ad sales sell the segments given to them - or sell the demographic details of their audience profiles to capture advertisers' budgets. Also they want volume of ad impressions, so even a large fat News segment sold Run of Site (ROS) is better than carving content up into many small segments. But, finding more "Technology" ad impressions within News, that enlarges the pool of "Technology" ad slots to sell, both inside and outside of the fixed editorial channel, starts to make a lot more sense.

The one problem with editorial segments is the story is often either in one category, or another category - never seemingly in both categories at the same time.

Grapeshot equips ad sales to look at the keywords in the article and not just the general assignment of a story to an editorial bucket of "News", so that more premium ad impressions or new unique channels can be unlocked for the benefit of the advertiser, even in general "News".

With Unique Tenancy channels advertisers can 'own' a series of keywords or key concepts. For example, imagine the power of creating a bespoke and unique 'Honda' channel for the launch of their hybrid. Honda could select and own the environment word cloud containing an almost infinite amount of keywords defining the concept.

Once created this unique target audience would be deployed where any page view throws up either the optimum score on content match or indeed the optimum score on a user based on their interests. This gives the advertiser the perfect campaign as it is delivered on evidence of overlapping keywords, and not supposition.

Grapeshot can look at all the words in an article and start to match the essence of the page - using the group of most significant words - to determine which categories overlap the best. Categories can be defined using keywords too, and the result is several categories can score an overlap. It means stories about "wind turbines" can match both an "Environment" channel, as well as a "Technology" channel, as well, perhaps, as a "Business" channel.

The other problem with editorial segments is they are often binary. An article is either "Environment" or not, "Business" or not. Grapeshot instead ranks categories against an article, so that there is a score of relevance, or of confidence, about the overlap.

Grapeshot provides a range of matching confidence, so we can assess which of several categories are relevant to the one story inside "News".

Now the ad sales team has several categories for a single story and can start to be flexible about how channel categories are sold and trafficked through the ad server.

Using percentages and scores, Grapeshot can offer the ad sales team the opportunity to throttle one category or channel up against another. Both channel categories can match an article - so for example the "wind turbine" story can be both "Technology" and "Environment" - but ad sales can determine the priorities of the match.

Why? Because there are both "Environment" and "Technology" channel ad campaigns that need to be sold and delivered for that month, and they can now start to manipulate which one gets priority to meet the monthly target numbers!

Why wait for the vagaries of what each journalist writes and the frequency at which visitors alight and read content. Instead the ad sales team can control the very channel categories that get sent downstream to the DoubleClick (or any other) ad server.

No more do categories have to be driven by manual tags inserted by hand within the ad call, or automatically lifted from the Editorial section of the website, or manually assigned to URLs as part of a large behavioural network. Instead, Grapeshot assigns a virtual category in real-time, to instruct the ad server on the context of the page or the context of the user viewing that actual page view - in real time.

The real-time nature of Grapeshot channel categorisation, means ad sales can develop new categories, adjust old ones and control with the Grapeshot Throttle the volumes of channel categories being sent to the ad server. Now they do have the control to manage the ad impressions they sell, from creation through to execution. Ad sales have the control to find the best 100,000 ad impressions for the 100,000 ads booked, for each and every channel campaign.

The fact that ad wastage is reduced, and click though rates rise - is just a by product of delivering relevant ads evidenced by the words on the page rather than relying on the editorial section in which a journalist happens to placetheir original story. Finding "technology" channel articles inside "News" sections can also improve the premium price points at which ad campaigns are sold. Always evidenced by the actual words that created the keyword matches, this virtual channel method is in fact more robust than the traditional manual process that can often misplace stories in the wrong premium sections.