The usual way of choosing a clients management system is to look at products and react to what they show you, which means your requirements are written by vendors. The alternative takes an hour and changes the outcome more than any amount of comparison. Write one page first: what you keep records about, the specific questions you cannot answer today, what has to survive from your current setup, and who will maintain the thing in eighteen months. Then let products fail against that page rather than impress you against nothing.
Name the unit of work, and your two awkward cases
Write down what you keep records about: an individual, a household, a company with several contacts, a matter. Then name the two client types that are awkward, because every product handles the straightforward case and they differ entirely on the edges. Those two cases are your evaluation. Reproduce them in every trial and eliminate anything that needs a workaround, because that workaround becomes your daily routine.
List the questions you cannot answer today
Which clients have not been contacted in six months. What is outstanding from whom. Which clients are on which service. Who owns this relationship. Most practices can name five such questions in a couple of minutes, and those five are the actual value of the purchase. A product that answers them is worth buying; one that merely stores your data more attractively is not.
Decide who maintains it, and what must come across
Name the person who will add clients, retire old ones and adjust the setup. If they are not technical, usability is a requirement rather than a preference. Then list what must survive the move: client details, documents with their client and date, and any history you would miss. Anything not on that list can stay in an archived copy of the old system, and saying so now prevents a migration that tries to bring everything.
Questions people ask about clients management system
How long should choosing a clients management system take?
An hour writing the page, then two weeks trialling the two products that survive it. Longer evaluations mostly produce more comparison, not better information.
Should we involve staff in the choice?
Involve whoever will use it daily, especially in the trial. The person who will maintain the system has a veto worth respecting, because they inherit every compromise.
What if no product matches the requirements page?
Then relax the least important requirement rather than adding a second system. Two systems is the outcome that reliably costs more than any single imperfect one.