WEBVTT

00:00:00.000 --> 00:00:05.860
Hey everybody and welcome back to the Operational ITAM Podcast. I'm Bill Van Nort, and today, we

00:00:05.860 --> 00:00:10.990
are reading a meter. On July 16th, a New York fintech company called Ramp

00:00:11.310 --> 00:00:15.370
put out a press release with a number in it that I have been thinking about ever since.

00:00:16.120 --> 00:00:20.450
Ramp processes corporate card and bill payments for more than 70,000 American

00:00:20.450 --> 00:00:23.470
businesses, which means they are not surveying anybody.

00:00:24.030 --> 00:00:27.610
They are watching the money move. And what they reported it is this.

00:00:29.000 --> 00:00:34.990
Since June of 2025, AI token spend across their customers has increased 20.7

00:00:34.990 --> 00:00:40.680
times, not 20%, 20.7 times in about 13 months.

00:00:41.310 --> 00:00:46.080
That is not a line item growing. That is a new category of spend arriving at

00:00:46.080 --> 00:00:51.110
full speed, and it does not behave like anything your procurement process was built to approve.

00:00:52.400 --> 00:00:56.730
Nobody signed a purchase order for that number. It accrued. One request at a

00:00:56.730 --> 00:01:00.161
time on a meter almost nobody in the organization has ever read.

00:01:13.000 --> 00:01:13.337
Good morning, good afternoon, or good evening.

00:01:17.200 --> 00:01:22.830
This is the show where we take the unglamorous machinery of enterprise technology and make it make sense.

00:01:23.330 --> 00:01:28.860
Grab your coffee. Last episode, we did something we had never done before. We closed the library.

00:01:29.530 --> 00:01:34.170
11 episodes in those rooms, and we walked out with full honors and stepped into

00:01:34.170 --> 00:01:38.560
the grid because the estate we managed stopped behaving like a collection and

00:01:38.560 --> 00:01:40.110
started behaving like a utility.

00:01:40.910 --> 00:01:44.820
Today is the first stop on the tour I promised you, the meter room.

00:01:45.560 --> 00:01:49.340
In episode 11, we laid out the five forms AI takes in your estate,

00:01:49.810 --> 00:01:54.320
and one of those forms was metered, the form where spend scales with usage.

00:01:54.890 --> 00:01:59.840
Today, we walk all the way into that one. And honestly, the warning shot came even earlier.

00:02:00.420 --> 00:02:07.120
Back in episode 9, Zylo's SaaS Management Index told us 78% of IT leaders got

00:02:07.120 --> 00:02:10.250
hit with unexpected consumption or AI charges.

00:02:10.900 --> 00:02:15.570
At the time, that was a renewals problem. It is not a renewals problem anymore.

00:02:16.040 --> 00:02:17.770
It is the shape of the whole market.

00:02:18.610 --> 00:02:22.889
So let me give you the numbers and I'll name the sources because on this show, we do that.

00:02:23.740 --> 00:02:28.810
Ramp publishes benchmarks from their token spend product, real payments, not survey answers.

00:02:29.120 --> 00:02:32.850
And the April numbers tell you everything about how this category behaves.

00:02:33.500 --> 00:02:37.960
The median business paid about $2,246 a month for AI.

00:02:38.430 --> 00:02:42.590
The average paid about $140,842.

00:02:43.949 --> 00:02:49.660
Sit with that gap for a second. When the average is 60 times the median, that is not a typo.

00:02:50.260 --> 00:02:54.430
That is a distribution with a long, expensive tail. And the distance between

00:02:54.430 --> 00:03:00.650
typical and tail is usually one team or one automated workflow running without a ceiling.

00:03:01.290 --> 00:03:06.350
Same dataset. The effective rate across everything averaged $0.72 per million

00:03:06.350 --> 00:03:12.590
tokens, but $1,000 buys you roughly 14.5 billion tokens on the cheapest mainstream

00:03:12.590 --> 00:03:15.790
model and about $704 million on a flagship.

00:03:16.500 --> 00:03:21.820
Same money, 20 times difference, decided by a default someone set in an afternoon.

00:03:22.450 --> 00:03:26.720
And the FinOps Foundation, in their sixth annual State of FinOps survey,

00:03:27.080 --> 00:03:32.380
nearly 1,200 practitioners found that 98% now manage AI spend.

00:03:32.980 --> 00:03:35.440
Two years ago, that was 31%.

00:03:36.060 --> 00:03:41.690
The single most requested capability in the entire survey was granular AI spend monitoring.

00:03:42.180 --> 00:03:45.410
Tokens, requests, GPU utilization.

00:03:45.940 --> 00:03:50.000
The people whose whole job is reading bills are telling us the bill they cannot read yet.

00:03:51.160 --> 00:03:55.430
So here is the thesis. The price of software used to be set at the signature.

00:03:55.820 --> 00:03:57.330
Now it accrues at the meter.

00:03:57.880 --> 00:04:02.400
A seat is a price you approved once. A meter is a price you discover every month.

00:04:02.940 --> 00:04:06.420
And the discipline this demands is simple to say and rare to find.

00:04:06.930 --> 00:04:08.870
Read the meter before the invoice does.

00:04:09.720 --> 00:04:12.730
So let's open the panel and look at what a meter actually is.

00:04:13.090 --> 00:04:17.000
Because every consumption agreement in your estate, AI or otherwise,

00:04:17.290 --> 00:04:23.080
is built from the same four parts, and you need to be able to name all four for every meter you own.

00:04:23.899 --> 00:04:27.550
Part one is the unit. What makes the meter click forward by one?

00:04:28.170 --> 00:04:32.180
For a language model, it is the token, and here is the trap inside the trap.

00:04:32.870 --> 00:04:36.110
Tokens in and tokens out are different units at different prices,

00:04:36.460 --> 00:04:39.770
and output typically runs three to five times the cost of input.

00:04:40.460 --> 00:04:44.120
For infrastructure, it might be the GPU hour. For a platform agent,

00:04:44.120 --> 00:04:45.520
it might be the message, or the

00:04:45.520 --> 00:04:49.670
run, or a credit, which is a currency the vendor invented and controls.

00:04:50.320 --> 00:04:53.010
If you cannot say what the unit is, stop.

00:04:53.720 --> 00:04:56.660
Everything else about the meter is unknowable until you can.

00:04:57.580 --> 00:05:01.960
Part two is the rate, and this is where consumption pricing gets genuinely strange

00:05:02.300 --> 00:05:04.870
because the same unit does not have one price.

00:05:05.620 --> 00:05:09.029
The same million tokens can bill at the full rate, at a cached rate,

00:05:09.410 --> 00:05:12.149
or at a batch rate on the same meter in the same month.

00:05:13.190 --> 00:05:17.450
RAMP's April data makes this concrete. Businesses running one popular mid-tier

00:05:17.450 --> 00:05:22.770
model paid an effective $0.62 per million tokens against a list price of $3.

00:05:23.390 --> 00:05:27.140
That entire gap is caching. Same model, same work.

00:05:27.810 --> 00:05:32.700
Fifth of the price because somebody engineered the request to reuse what was already paid for.

00:05:33.410 --> 00:05:36.190
When your effective rate and the list price diverge that much,

00:05:36.620 --> 00:05:38.399
the list price stops being information.

00:05:38.920 --> 00:05:43.560
Your effective rate, your dollars divided by your units, is the only rate that is real.

00:05:44.660 --> 00:05:48.850
Part 3 is the commitment. Most enterprise consumption deals have a floor.

00:05:49.460 --> 00:05:52.750
You promise to burn a number, and the meter runs against that promise.

00:05:53.339 --> 00:05:57.610
Which means a meter can hurt you in both directions. Overage above the commitment,

00:05:57.940 --> 00:06:01.529
or a commitment you never consume, which is shelfware with a clock on it.

00:06:02.399 --> 00:06:05.790
Reading the meter mid-cycle means comparing the accrual against the commitment

00:06:06.040 --> 00:06:07.450
while there is still time to steer.

00:06:08.770 --> 00:06:12.760
And I will hand you some humility on this one. Mine. Years ago,

00:06:12.760 --> 00:06:16.730
on a purchase order I signed, I sized a consumption commitment straight off

00:06:16.730 --> 00:06:18.380
the vendor's projection of our growth,

00:06:18.920 --> 00:06:24.050
because the discount at that tier was beautiful, and the projection came with a very confident chart.

00:06:24.800 --> 00:06:27.820
We never read the meter mid-cycle. Not once.

00:06:28.390 --> 00:06:31.080
The true-up arrived at the end of the year like a weather event,

00:06:31.500 --> 00:06:33.640
and I remember thinking we had been surprised.

00:06:34.940 --> 00:06:39.640
We had not been surprised. We had been unread. The meter knew the whole time.

00:06:40.500 --> 00:06:43.940
So let me give you an opinion here, clearly labeled as an opinion.

00:06:44.390 --> 00:06:47.910
Size a consumption commitment to what your own history says you will actually

00:06:47.910 --> 00:06:51.620
burn. Your typical month. Not the vendor's growth curve for you.

00:06:52.290 --> 00:06:54.970
The counter-argument is real, and I'll give it to you fairly.

00:06:55.620 --> 00:06:57.280
Bigger commitments buy better rates.

00:06:57.630 --> 00:07:01.840
And if your usage genuinely is doubling, committing low means paying overage

00:07:01.840 --> 00:07:03.290
on growth you should have seen coming.

00:07:03.840 --> 00:07:08.030
Fine, then let your own meter readings make that case, three months of them,

00:07:08.360 --> 00:07:10.920
and commit on your evidence instead of their enthusiasm.

00:07:11.850 --> 00:07:16.250
And part four is the true up. What happens when actual crosses committed?

00:07:16.870 --> 00:07:18.480
What rate applies above the line?

00:07:19.130 --> 00:07:21.530
When does the excess bill and can it carry over?

00:07:22.270 --> 00:07:25.540
Those four answers live in the contract and I will tell you from experience,

00:07:25.950 --> 00:07:30.140
almost nobody who watches the dashboard has read the contract and almost nobody

00:07:30.140 --> 00:07:32.150
who read the contract watches the dashboard.

00:07:32.840 --> 00:07:36.210
Now let me show you what this looks like when real organizations do it well,

00:07:36.630 --> 00:07:41.310
because two examples crossed the wire this summer and they teach opposite ends of the same lesson.

00:07:42.110 --> 00:07:46.770
First one. In late June, Coinbase's chief executive, Brian Armstrong,

00:07:47.050 --> 00:07:51.240
posted their internal playbook, and the trade press picked it apart in detail.

00:07:51.960 --> 00:07:56.600
Coinbase cut its AI spend nearly in half while its token usage kept growing.

00:07:57.830 --> 00:08:03.880
Read that again. Half the spend. More usage. And the part I want you to hear is how.

00:08:04.530 --> 00:08:10.000
Not caps. Armstrong said 91% of employees never even hit the old usage caps,

00:08:10.370 --> 00:08:12.810
so the caps were friction pretending to be control.

00:08:13.440 --> 00:08:18.120
His words on how you actually do it, quote, not with friction and spend alerts,

00:08:18.470 --> 00:08:21.890
with better defaults, routing, and caching, end quote.

00:08:22.580 --> 00:08:25.910
Their gateway defaults route routine work to inexpensive models.

00:08:26.340 --> 00:08:32.470
Harder work escalates when it earns it, and their caching went from hitting 5% of requests to 60%.

00:08:33.309 --> 00:08:36.110
Every one of those decisions happened before the meter turned.

00:08:36.770 --> 00:08:39.020
Second example, from the other side of the house.

00:08:39.660 --> 00:08:43.160
In Ramp's launch announcement, the controller at AngelList describes getting

00:08:43.160 --> 00:08:46.470
a weekly spend briefing that surfaced a term he had never heard.

00:08:47.390 --> 00:08:51.970
Prompt caching. Not a controller's phrase. He routed it to engineering.

00:08:52.320 --> 00:08:57.080
They found they had been losing about $10,000 a month, and the fix went in the same day.

00:08:58.030 --> 00:09:01.700
Think about what actually happened there. The control lived in engineering.

00:09:02.080 --> 00:09:03.790
The visibility lived in finance.

00:09:04.260 --> 00:09:06.820
The money got saved on the day those two met in the middle.

00:09:07.429 --> 00:09:12.380
That meeting place is the meter room, and in most organizations, that room is empty.

00:09:13.320 --> 00:09:16.360
Let's take a quick break. If you're getting value from this show,

00:09:16.520 --> 00:09:18.000
subscribe wherever you're listening.

00:09:18.540 --> 00:09:22.220
And if you know somebody who just opened an AI invoice they could not explain,

00:09:22.520 --> 00:09:25.660
send them this episode because the explanation is in part 2. Every episode with full transcripts is at operationalitam.com.

00:09:31.929 --> 00:09:35.309
Ok, Part 2 – How you actually read a meter before the invoice does.

00:09:35.750 --> 00:09:37.960
Step 1 – Find every meter.

00:09:38.640 --> 00:09:41.740
A meter you have not located is a meter the vendor reads alone,

00:09:42.130 --> 00:09:45.350
and your meters are hiding in all five forms from episode 11.

00:09:46.070 --> 00:09:50.420
If you did that episode's homework, the Shadow AI Census, you already hold the

00:09:50.420 --> 00:09:55.480
beginning of your meter list – expense data, single sign-on, asset records.

00:09:56.090 --> 00:09:59.670
Every AI vendor on that census either bills a seat or bills a meter,

00:10:00.049 --> 00:10:02.870
and your first pass is just sorting them into those two piles.

00:10:03.590 --> 00:10:06.880
Add the cloud AI services riding inside your existing cloud bill,

00:10:07.230 --> 00:10:10.130
because those meters do not even arrive as separate invoices.

00:10:10.530 --> 00:10:15.120
They arrive as line items inside a bill you already pay, under your cloud provider's

00:10:15.120 --> 00:10:18.190
name, which is exactly why the census keeps missing them.

00:10:18.850 --> 00:10:25.400
Step 2. Reconcile every invoice line to a meter you found. And I mean arithmetic, not vibes.

00:10:25.920 --> 00:10:30.030
Take the dollars on the line, divide by the units consumed, and write down the

00:10:30.030 --> 00:10:32.320
effective rate, then do last month.

00:10:32.910 --> 00:10:36.390
A line you cannot tie to a meter is not a cost you approved,

00:10:36.700 --> 00:10:40.929
it is a cost you absorbed, and an effective rate that moved is telling you one

00:10:40.929 --> 00:10:42.690
of exactly three stories.

00:10:43.460 --> 00:10:47.150
The mix changed, meaning somebody moved work to a pricier model.

00:10:47.770 --> 00:10:51.320
The caching changed, meaning the engineering got better or worse.

00:10:51.799 --> 00:10:54.740
Or the price changed, meaning the vendor moved the rate.

00:10:55.360 --> 00:10:59.299
Three stories, three different owners, and the invoice total tells you none

00:10:59.299 --> 00:11:02.049
of them. The effective rate tells you all of them.

00:11:02.790 --> 00:11:07.630
Step 3. Watch the drivers, not the total. Ramp's data says the single biggest

00:11:07.630 --> 00:11:12.150
driver of surprise cost is model tier migration, teams upgrading to premium

00:11:12.150 --> 00:11:14.400
models for quality, usually invisibly.

00:11:15.059 --> 00:11:24.520
Premium models went from 5.7% of business token cost in June of 2025 to 55.9% by April, 10 months.

00:11:25.100 --> 00:11:28.650
And the difference per workflow can run 10 to 100 times.

00:11:29.510 --> 00:11:32.350
The other driver is agents, and I want to say this one carefully,

00:11:32.700 --> 00:11:34.760
because it is the future of this problem.

00:11:35.470 --> 00:11:40.100
An agent takes as many steps as it decides the task needs, which means you approved

00:11:40.100 --> 00:11:42.690
the task, and the agent is deciding the bill.

00:11:43.350 --> 00:11:48.179
Any automated workflow without a per-run ceiling is a meter with no shutoff valve.

00:11:48.920 --> 00:11:53.070
Step 4. Set the defaults, because this is the actual governance layer.

00:11:53.670 --> 00:11:57.880
The Coinbase lesson is not that they saved money, it is where the saving lived.

00:11:58.400 --> 00:12:02.130
In the gateway default, in the routing rule, in the cache policy.

00:12:02.809 --> 00:12:07.110
Decisions made once, upstream, that shape every request afterward.

00:12:07.760 --> 00:12:12.790
A spend alert fires after the money is gone. A default decides whether it ever leaves.

00:12:13.740 --> 00:12:17.679
And one thing I am deliberately not giving you today, there is a token price

00:12:17.679 --> 00:12:21.340
decline statistic making the rounds, a big satisfying percentage.

00:12:21.910 --> 00:12:25.700
I chased it to its source, the way we did with refresh numbers back in episode

00:12:25.700 --> 00:12:29.960
7, and it traces to a platform vendor measuring its own customer base.

00:12:30.450 --> 00:12:31.710
So you will not hear it here.

00:12:32.429 --> 00:12:35.520
What I can give you is Ramp's paired numbers across the same window.

00:12:36.120 --> 00:12:43.110
Token usage among businesses up 1,001% January 2025 to April 2026.

00:12:43.720 --> 00:12:45.840
Spend up 497%.

00:12:46.630 --> 00:12:50.870
Falling unit prices are real, and they are not saving anybody money because

00:12:50.870 --> 00:12:54.820
usage is growing faster than prices falling and it is not close.

00:12:55.860 --> 00:13:00.760
Which brings us to a listener question, and this one landed in my inbox from Elena in Cincinnati.

00:13:01.470 --> 00:13:06.650
Elena writes, Our finance team wants next year's AI budget as one fixed number.

00:13:07.200 --> 00:13:10.980
With consumption pricing, that feels impossible. What do I even give them?

00:13:12.000 --> 00:13:16.460
Elena, your instinct is right, and here is the move. Do not give them one number,

00:13:16.800 --> 00:13:21.800
because one number is a guess dressed up as a commitment, and finance people can smell that.

00:13:22.350 --> 00:13:25.920
Give them three numbers, reported separately, never blended.

00:13:26.640 --> 00:13:32.190
Number one, the committed floor. That is contractual. It is the most honest number you own.

00:13:32.770 --> 00:13:36.950
Number two, the expected run rate. Built from this year's effective rates and

00:13:36.950 --> 00:13:41.510
named drivers and put a real buffer on it because ramps data shows month-over-month

00:13:41.510 --> 00:13:45.380
swings above 40% are common even with stable headcount.

00:13:45.980 --> 00:13:49.660
When finance asks why the buffer is that big, do not defend it.

00:13:50.140 --> 00:13:53.060
Show them the swing data and let the data do the arguing.

00:13:53.860 --> 00:13:58.460
Number three, the uncapped exposure. And next to every meter that creates it,

00:13:58.780 --> 00:14:00.550
the specific control that caps it.

00:14:01.160 --> 00:14:04.770
The per run ceiling, the gateway default, the alert threshold.

00:14:05.480 --> 00:14:09.570
Here's the reframe, Elena. You cannot promise finance what a meter will read,

00:14:09.830 --> 00:14:11.530
and you should stop pretending you can.

00:14:12.180 --> 00:14:15.880
What you can promise is which drivers are governed and who owns each one.

00:14:16.620 --> 00:14:20.760
Budget the drivers. Report the meter. That conversation is the whole job.

00:14:22.000 --> 00:14:25.510
All right, the grid room. We are standing in the meter room this week,

00:14:25.920 --> 00:14:29.410
and if you have ever actually looked at the utility meters in a commercial building,

00:14:29.790 --> 00:14:33.530
you know three things about them that map onto everything we just covered.

00:14:34.050 --> 00:14:36.630
First, the meter spins whether anyone reads it.

00:14:37.490 --> 00:14:41.920
It does not care about your budget cycle, your approval workflow, or your fiscal year.

00:14:42.500 --> 00:14:46.440
Consumption is recorded the moment it happens, and the only question is who

00:14:46.440 --> 00:14:48.790
reads it first, you or the biller.

00:14:49.850 --> 00:14:53.000
Second, utilities have a thing called an estimated read.

00:14:53.560 --> 00:14:57.320
When nobody physically reads the meter, the utility guesses your usage from

00:14:57.320 --> 00:15:00.690
history and bills the guess, and they reconcile it later.

00:15:01.410 --> 00:15:04.220
And I want you to notice whose model produces the guess.

00:15:04.990 --> 00:15:09.420
When you do not read your own meter, you are living inside the vendor's estimate of you.

00:15:10.700 --> 00:15:15.990
Third, and this is my favorite, commercial power bills carry something called a demand charge.

00:15:16.650 --> 00:15:20.900
You do not just pay for the energy you used. You pay a premium pegged to your

00:15:20.900 --> 00:15:25.280
single worst 15 minute window of the month. Your peak sets the price.

00:15:26.140 --> 00:15:30.280
Now think about an agent workflow with no ceiling, spiking on a Friday night.

00:15:30.960 --> 00:15:36.480
Consumption billing is demand charge economics, and your estate is full of peaks nobody is watching.

00:15:37.340 --> 00:15:41.390
The utilities figured all this out a century ago. They read on a schedule.

00:15:41.710 --> 00:15:44.700
They audit the estimates. They engineer the peaks down.

00:15:45.330 --> 00:15:48.100
None of it is new. It is just new to us.

00:15:49.240 --> 00:15:53.620
Which brings us to today's principle. And let's take the whole run from the top.

00:15:54.210 --> 00:15:59.010
Hardware asset management is a custody discipline. Software asset management

00:15:59.010 --> 00:16:00.320
is an evidence discipline.

00:16:01.030 --> 00:16:05.660
Audit defense is a process discipline. Settlement is a commercial discipline.

00:16:06.340 --> 00:16:08.610
Shadow IT is a service discipline.

00:16:09.300 --> 00:16:11.340
Refresh is an economics discipline.

00:16:12.030 --> 00:16:16.470
Disposal is a liability discipline. Renewal is a leverage discipline.

00:16:17.150 --> 00:16:21.730
Tooling is a judgment discipline. The AI estate is a governance discipline.

00:16:22.350 --> 00:16:25.570
And consumption? Consumption is a measurement discipline.

00:16:26.260 --> 00:16:30.080
The seat asks you to count people. The meter asks you to measure flow.

00:16:30.690 --> 00:16:35.060
And a flow you do not measure is a price you do not know, on a bill you have

00:16:35.060 --> 00:16:37.820
already agreed to pay. Class dismissed.

00:16:39.080 --> 00:16:44.000
Here is your homework and it needs about an hour. Pull your largest AI or consumption

00:16:44.000 --> 00:16:45.590
invoice, last full month.

00:16:46.350 --> 00:16:51.570
Go line by line and do two things for every line. First, name the meter behind

00:16:51.570 --> 00:16:53.590
it, the unit and the rate.

00:16:54.280 --> 00:16:58.700
Second, divide the dollars by the units and write down the effective rate,

00:16:59.110 --> 00:17:00.900
then do the same for the month before.

00:17:01.800 --> 00:17:05.760
When you are done, you will hold a one-page list of two kinds of findings,

00:17:06.280 --> 00:17:10.710
lines with no meter you can name, and rate movements with no story you can tell.

00:17:11.470 --> 00:17:15.880
Date the page, sign it. That page is your first meter reading,

00:17:16.200 --> 00:17:19.890
and whoever you show it to will understand this episode without hearing a word of it.

00:17:21.160 --> 00:17:24.740
Next episode, we walk from the meter room to the standards body,

00:17:25.310 --> 00:17:28.530
because here is what happened while nobody in ITAM was watching.

00:17:29.359 --> 00:17:34.180
The FinOps Foundation built a common language for exactly the bills we struggled with today.

00:17:34.890 --> 00:17:39.520
It is called FOCUS, the FinOps Open Cost and Usage Specification.

00:17:40.170 --> 00:17:43.790
And this June, their steering committee ratified version 1.4,

00:17:44.160 --> 00:17:48.930
which adds invoice reconciliation, the exact discipline we just did by hand,

00:17:49.290 --> 00:17:54.300
straight into the standard, with native AI support already on the roadmap for the next version.

00:17:55.140 --> 00:17:58.880
These are the people who renamed their own flagship conference around the token.

00:17:59.760 --> 00:18:04.030
The meter readers built themselves a language. Next week, we learn to speak

00:18:04.030 --> 00:18:08.750
it, and I will show you what it means for how ITAM and FinOps stop working the

00:18:08.750 --> 00:18:10.560
same estate in different rooms.

00:18:11.700 --> 00:18:16.109
And listen, the case files are open. One situation, one page.

00:18:16.570 --> 00:18:21.109
The constraint you were under, what you did, what happened. No company names,

00:18:21.109 --> 00:18:22.520
I will sanitize everything.

00:18:22.920 --> 00:18:26.753
Send me one worth working at the website and I will build an episode around it. I'm Bill Van Nort.

00:18:26.753 --> 00:18:26.806
This is the Operational ITAM Podcast. Find every meter. Compute your own effective rate.

00:18:26.806 --> 00:18:26.859
Set the default before the invoice sets it for you. I'll talk to you next week. Take care.