The data doesn't seem to add up to me and more context may be needed, it will help to explain what the exclusive and duplicate reach are. A diagram of what the venn diagram should look like would help too and which column and cells equate which overlaps.

My suspicion though is that you may be missing the data to make a Venn view in the first place. You basically will need some cells saying:

Site A only, Site A+B, Site A+C,
Site B only, Site B+C,
Site C only
Site A + B + C ( this one is missing)

Also the Site A/B/C only ones are dubious as I don't think the exclusive reach equates those.

Hi Daniel, I have attached another one that has more sites and I've left the formulas used to calculate the overlap. Regarding the sums, I have the impressions per site, so do you think I would need to create columns (field) that would show the sum of those? Cheers,

Unfortunately what you have here isn't possible to recreate in a Venn diagram, firstly because you have 5+ sites which is the maximum the Venn diagram can support in Omniscope. From your last comment, it seems to suggest you will have cases with even more sites so there lies more problems.

Secondly and more important is that the overlapping information you have is too superficial and lacks the detail in multiple overlaps i.e. which overlaps is A+B+C+D or A+C+E etc. This detail is absolutely necessary for Venn to work as currently you have no way of working out how all the site circles will overlap each other and by how much.

I would suggest you try using the radial view instead if this is the extent of data available. I can see how it may be difficult to get the necessary information as what you have resembles the basis of a complicated set of spatial interaction calculations.

Without actually knowing how the calculations are done for the numbers I don't think I can recreate this any better than what is achieved with the radial view.

There are several issues that are characteristic for the Venn View - it is informative and adds value when number of subsets shown is 2-4, and not more than 5-6.
Venn uses queries to define subsets and works well with granular data, where individual records can be grouped according to value in different fields e.g. gender=male, or age=40. This will allow single record to be found in results of two or more different queries, creating the 'overlap' between the subsets.
Problem with your data set was the structure - data was already aggregated, and organised as cross-table, therefore not ideal material for the Venn analysis.
Data orientation can be fixed in Omniscope using Pivot and De-pivot operations (please see attached file).
Further to that, I re-named some values and tokenised the result field.
Depending on the 'big picture', how many sites you are looking at, other views might offer better visualisation, while Venn is always the most effective when you are focusing on just 2-3 at a time. Having said that, you could analyse 100 sites and define 100 queries (very quick, using 'Create query from filter'), and select 2-3 at a time, for close-up analysis.
Note -Knowledge of the data set is necessary to understand the split in data categories, and whether overlap between sites is already part of "Duplicates".

Thank you Paola, that solution works perfectly.