<aside> ❓

The questions I asked of every app in the benchmark. Asking each app the same questions kept the comparison fair and made the gaps easy to see.

</aside>

Problem area 1: Feature overload

Goal: find opportunities for AI-powered personalisation and adaptive experiences.

#	Question	Good sign	Warning sign
1	How are features prioritised or surfaced?	Home leads with what matters now: upcoming payment, spending summary, cards	Cards, crypto, insurance, investments, loans and cashback all compete equally
2	How easy is it to find the relevant action on the home screen?	"Your subscription renews tomorrow" is visible straight away	Analytics → Payments → Recurring → Subscription
3	Does the experience adapt to behaviour?	A frequent traveller sees FX and travel spending	Everyone sees the same dashboard
4	What creates cognitive load or confusion?	Clear hierarchy and labels	Too many options, unclear labels, dense dashboard
5	Where is the opportunity to simplify?	Unused features recede, relevant actions come forward	A static interface for every user

Problem area 2: Subscription creep

Goal: find opportunities for predictive subscription support.

#	Question	Good sign	Warning sign
1	How are recurring payments detected and shown?	"Spotify detected as a monthly payment"	The user tags subscriptions manually
2	Is support proactive or reactive?	"Netflix renews in 3 days"	"You spent 15 AZN on Netflix"
3	Does it help decide, or only inform?	"Renews in 3 days and you haven't used it recently. Consider cancelling."	"Your subscription renews in 3 days."
4	How much manual effort does managing subscriptions take?	One-tap review and cancel	Search transactions, identify the subscription, manage it elsewhere
5	Where is the opportunity to intervene earlier?	Likely unwanted renewals are flagged before they happen	Tracking only, after the charge

Goal: find opportunities for accurate categories, transparent insights and more trust.

#	Question	Good sign	Warning sign
1	How accurate are the categories?	Uber → Transport	Uber → Shopping
2	How easy is it to review and correct a category?	One-tap correction, and the app remembers it	The same merchant is fixed by hand every month
3	How transparent is the logic?	"Categorised as Food based on your previous spending"	A category with no explanation
4	How do errors affect trust in insights?	Consistent categories, trusted breakdowns	Wrong category → inaccurate report → low trust in advice
5	Where is the opportunity to improve accuracy and trust?	The system learns from corrections	Static rules that never improve for the individual

Goal: find where AI can be adaptive and trustworthy while reducing effort.

#	Question	Good sign	Warning sign
1	Does AI reduce effort and support decisions, or add more information?	"Renews soon and looks unused. Consider cancelling."	"Renews in 3 days."
2	Does AI learn from behaviour?	Category preferences are learned over time	The same recommendations for everyone
3	Does AI explain itself?	"Categorised as Transport based on your previous corrections."	The category changes with no explanation