How many tools is your company using? Maybe you know the answer to that. The hardest question is: how many tools is your company paying for? A marketing tech stack audit is a structured review of every tool you pay for: what it costs, who owns it, what job it does, and whether its data reaches the place you report from. Work through eight steps, score each tool out of five, then decide whether to keep it, connect it, consolidate it, or cut it.
Most audits fail at step one because they start with a list somebody typed from memory. That list is always shorter than the real one. The tools you forget are the tools nobody owns, and the tools nobody owns are the ones quietly charging you every month for a job something else is already doing.
So this guide starts with your bank statement.
The problem is not the number of tools. It's the half you don't use.
The surveys that track this have watched the same uncomfortable number for six years: the ratio of how much of the stack marketers use versus how much they bought. It ran at 58% in 2020, slid to 42% two years later, and hit a third by 2023 (1). The most recent read puts it back at 49%, with the marketers in it planning to reach 61% (2).
Sit with the 61% for a second. Even the optimistic plan, made by the people who bought the tools, leaves four of the ten tools idle. Nobody is using what they own.
The market keeps making it easy. There are now 15,505 marketing technology products on the market, and 1,488 of them arrived in the last year alone (3).
Scott Brinker, who has tracked that landscape since 2011, points to the fix most teams skip: "Probably the most underutilized remedy for underutilization is investing in good enablement" (4).
Hold onto that before you start canceling subscriptions. The money you recover from a marketing tech stack audit is real, and it is the smaller half of the prize. The larger half is the tool you already own, already pay for, and have never switched on.
What you need before you start
Three things, and about half a day.
- The last twelve months of card and bank statements for whatever card the software sits on. Not a list from memory.
- Read access to your CRM, your email platform, and your analytics.
- One person with the authority to cancel something.
That last one is not a joke. Audits that end in a document rather than a decision are why most stacks get audited twice.
Pull the real list from your statements
Open twelve months of statements and write down every charge that is for software. Twelve months, not one, because annual renewals are hidden from a monthly view and are usually the more expensive ones.
Add anything charged to a personal card and expensed.
Bonus points if you can ask the team to add any free tiers too: a free tool still holds customer data, still needs an owner, and still counts as a place your information lives. Maybe ask the team to search their inboxes for any "Welcome to..." or even "login" emails.
You will find things nobody remembers buying. That is the point of doing it this way.
Write down what each one costs per year
Convert everything to an annual number. A $79 monthly seat reads as small. The same tool reads as $948 a year, and four of those read as a part-time hire.
Note the renewal date next to each. You cannot cancel most things on the day you decide to. You can only cancel them within a window, and that window closes before the date you didn't write down.
Give every tool one named owner
One human name per tool. Not a team, not a role, a person.
If nobody's name goes in the box, you have found your first finding. An unowned tool isn't underused; it is being left running.
Map each tool to the one job it does
Write the job in plain language, the way you would say it out loud: "This is where email lists live," "this is where we book calls," "this is where we see what pages people read."
Then sort the list by job. The overlaps announce themselves. Most stacks carry two tools that both send email, two that both hold contacts, and one analytics tool nobody has opened since it was installed.
Some overlap is fine. Two email tools can both be right if one handles sales sequences and the other handles newsletters. Two email tools where nobody can say which owns which list is a data problem waiting to be found by a client who gets the same message twice.
Score every tool out of five
Five questions, one point each. Answer them with evidence rather than intent.
- Owned. Does one named person own it?
- Used. Did somebody log in this week?
- Singular. Is it the only tool doing that job?
- Connected. Does its data reach your reporting without a manual export?
- Missed. If it vanished on Monday, would something visible break?
A 5 is a keep. A 3 or 4 is usually not a tool problem; it is a missing owner or a missing connection, and both are cheaper to fix than to replace. A 0 to 2 is a decision you have been avoiding.
The fourth question is what separates a stack from a system. A tool whose data can only be exported manually is not connected to anything. It is a room with no hallway, and every report built from it is somebody's afternoon.
Trace the path a customer takes through the stack
This is where we start being strategic. Pick one real person who became a customer in the last quarter. Follow them.
Where did they first show up, and which tool recorded it? What happened when they filled in the form: which system received it, what fired next, and who got notified? When they booked, did the booking know they had been on your list for eight months? When you report on that deal, which tool gets the credit, and is it the right one?
One real person is enough to expose every break in the chain. This is the step people skip because it feels anecdotal. It finds what a spreadsheet cannot: the handoff that never happens, the field that never gets written, the notification that goes to someone who left.
In my opinion, the most important step. This is when you connect an operational audit to your actual customer journey and your own strategy.
Decide: keep, connect, consolidate, or cut
Four decisions, and every tool gets exactly one.
- Keep. Scored 5. Leave it alone and move on.
- Connect. Does a real job, owned, used, and tracked in your reporting. The fix is an integration or a named process, not a new purchase.
- Consolidate—two tools, one job. Pick the one with better data, move the work, and set a date to switch the other off.
- Cut. Nobody owns it, nobody logs in, nothing breaks. Cancel it and export the data before you say goodbye.
Write the decision next to the tool in the table below. A decision with no date next to it is an opinion.
Set the next audit date and a rule to protect it
Put the date in the calendar now. Twice a year for most teams, quarterly if you are adding tools faster than that (nudge to all marketing teams that are always testing the new). Write down what changed and what it saved, because the second amarketing tech stack audit takes a fraction of the time the first one did, but only if the first one left a record behind.
Then give the stack a rule that holds between audits, or you will do this same work again in six months.
Ours is one in, one out. We review software spend at our monthly financial review, and when a new tool comes in, we retire an old one that same month. It is not a clever rule. It works because it moves the decision to the moment of purchase, while somebody still remembers why they wanted the thing, instead of to an audit a year later when nobody does.
The honest objection: sometimes you need two new tools, and there is nothing worth cutting. Then break the rule out loud in the meeting. A rule you break on purpose once a year still beats no rule, because breaking it makes somebody say the reason out loud. Stacks do not drift on the purchases people had to justify.
The Makreting Tech stack audit checklist
Copy the table from this link and fill in one row per tool. It is the whole audit on one page. Add a row for anything that does not fit a category. A tool with no category usually has no job.
What a good audit returns
Fewer tools, sometimes. That is the outcome everybody expects and the least interesting one.
The two that matter more: a stack where every tool has a name attached to it, and a reporting path that no longer runs through somebody's manual export. Both come out of the same afternoon, and neither one requires you to buy anything.
If the audit shows more disconnections than duplications, the problem is not your stack. It is the system architecture underneath it, and that is a different piece of work. It also quietly caps growth, even though every tool on the list looks fine.
Walk the stack once, then decide once.
A marketing tech stack audit is not a project. It is half a day, a table, and one person who is allowed to cancel things.
Don't leave this session with just a document. The version that works ends with four columns: decisions, a calendar date, and a rule that holds the line until then.
If you would rather have someone walk it with you, schedule a complimentary call, and we will run the eight steps on your marketing tech stack together.
