Q: How do I build user personas when my audience is so diverse? I sometimes find users who are very unique from others. Do I just keep adding one for each?
Think of a T-shirt store.
They typically keep 5 sizes in stock: S, M, L, XL, XXL
The majority of shoppers are able to wear one of the 5 available sizes, even though the fit might not be accurate down to a cm.
On the other hand, for those shoppers with rare body dimensions (outliers), the store may not have anything suitable on the racks (yet).
Similarly, you need to identify broad user themes that allow you to cater to at least 80% of your primary target audience.
Remember: If you were to pick a specific individual, it'll be highly unlikely that they perfectly map to an existing persona. You should, however, still be able to find a close enough "fit".
As for the "unique" cases, you would only consider creating one for them when they aren't that unique anymore (i.e. you see a sizable audience for those traits).
How many personas are too many?
Of course, this depends on your product. But, generally, as soon as you exceed six, you'll find yourself running thin on bandwidth to reflect meaningfully mapped experiences for each on your product.
========
User personas are not always differentiated based on demographics & job titles.
Consider slicing your audience using dimensions that would warrant a different set of product attributes, features & messaging.
Ex: Let's consider a ride-hailing app.
A 40-year old German professional looking to go to work in the morning & a 25-year old Dutch worker making an evening commute back home have similar needs. They can be clubbed to one persona as their age or nationality doesn't necessitate product variance (apart from a language translation layer).
However, two American mothers - one traveling to the airport alone with 3 huge suitcases & the other taking her sole toddler to the park - have similar demographics but exhibit vastly different product needs (one needs a bigger car, the other needs a baby seat).
========
Don't develop personas based on averages.
Ex: Let's say you're planning to build a transport service and want to develop profiles based on the distance people commute everyday.
You do a survey with 10. Five say they travel 20 KMs/day. The other 5 mention 100 KMs/day.
Now, if you were to conclude that an average commuter travels 60 KM per and base a persona on that - well, you're actually not targeting anyone. Averages are evil that way.
As a Product Manager, you might be asked a lot of questions during an interview. One of them includes technical questions. Here are 4 types of technical questions that you might come across.