Poq's current model works well - they build apps for mid-size brands. They correctly perceive that this is hard to scale, so Jun and I came up with three ways we might make it scalable. (One thing we didn't consider was trying to sell a platform version of their product to larger brands - I just couldn't see how any but the very largest brands could ever need this as their app needs are relatively modest.)
First, they could set up a franchising business and sell not only technical tools for creating apps, but also the marketing and customer-service processes to go with these, so a freelancer or small agency could buy a whole product to resell to their customers. I suggested that this might be particularly interesting if they can internationalise the product so they can market to potential franchisees in lots of different countries.
Second, they could simply automate as much of their current agency model as possible - including training for new staff and detailed procedures for app builders to follow - and scale it in the way large consultancies do, like Accenture or PWC. They would need to accept that this will be a people-intensive business but it is perfectly possible to make it work.
Last, they could look at pivoting hard, using the current business as a cash cow while they discover what other types of problems their current customers have, perhaps in an adjacent area where their skills can be applied - for example, perhaps brands need apps for their staff to use at trade shows when talking to buyers, as a replacement for a pricing book.