# Representing unique data and its breakdown

I'm stuck with a problem that has me stumped.
Context: I'm working on a dashboard which gives details about the breakdown of a user's data.

Here's the basic workflow:
1. User Uploads data- Emails and Phone numbers.
2. We then match the emails and phone numbers against our database and show number of matches we found. These are "Actual Matches". An email and phone matched to 1 user makes 1 Actual record.
3. Then, we derive and determine a few things from the matched data. This could be from parter databases, etc. This leads to more matches. These are "Output Matches".
4. Finally, we offer an option to integrate with a 3rd party product which can further increase the matches.

Here's the solution I came up with:

Problem:
1. I'm unsure if this is understandable for the user. Especially the breakdowns- 100 Emails and 100 Phones only total to 100 Actual matches because they belong to the same record. That is, 1 Phone and 1 Email make 1 Actual record only if they belong to the same record. So, 100 Emails and 50 Phones would have only 50 as the highest possible Actual matches. A lot of users think that the Actual matches is a total of the two. How do I better represent this data?
2. I want to convey to the user that their "Records Uploaded" has gone through layers of matching and computations. To the users, the current view seems like separate entities. How do represent this better?

• I would consider using Venn diagrams, because in your case the data from `step N` are always a superset of the data from `step N-1`, if I understood correctly. This creates a visual representation that matches the flow of the computations - you have an initial set, which is then augmented with more stuff, and so on. Apr 22, 2019 at 7:43
• Assuming you're using email+phone as a unique way to identify a single user, what happens when 2 emails are mapped to the same phone or vice versa? What happens when a single user has 2 emails and 2 phones -- your logic would count this as 2 users?
– SNag
Apr 22, 2019 at 9:19
• @Ashwin Nepal I am a little confused. When users upload data (emails, phones), are they required to enter one email and one phone for each record? Can they enter a phone but not an email or vice versa? If no, then you will never end up with a 100 email and 50 phones uploaded. Not sure if i'm missing something here. Meaning that you always have the same number of emails and phones under Records uploaded section, right? Apr 22, 2019 at 15:07
• @Mo'ath Yes, there are cases where they may not upload either the phone number or the email so we wouldn't have the same number of emails and phones. Apr 25, 2019 at 9:33
• @SNag If a user has 2 emails and 2 phones then they'd be counted as two different users. And if 2 emails are mapped to the same phone, it'd be counted as a single user. Apr 25, 2019 at 9:34

# Try to sequence it like a story. Describe a process in as few words as possible.

Here's a very simplified version, but perhaps rather than have the look right away at the data at bottom, you could use some subtext under each step describe the sequence:

Instead of 'actual' try 'exact', and explain what that meant in that phase. Then show the difference between Exact and Output, which happens after connecting additional data sources.

The final step in the story is a what could be step. Here's your better world, if you adopt X

I've simplified the visual design in this quick sketch, to focus more on the writing.

I'm not sure if it works, but I hid the data sources to see if they get the benefits by reading at first.

Verbally. “We processed the uploaded data and found 100 connected emails and telephone numbers. Here are the results.”

And at the table, instead of “Emails:100, Phones: 100” → “connected emails and phones: 100”

How about a view similar to this: