Sub-Metering Installation Checklist: 7 Steps to Get It Right the First Time

Your electricity bill jumped 30% over two quarters. Nobody can explain which department or tenant is responsible. Then your VP says the word "sub-metering" in a meeting and looks at you—the office administrator who happens to own the utility invoices.

That was basically my situation about three years ago. Sub-metering itself isn't complicated. The way it gets sold is. What follows is the checklist I now use for every energy monitoring project: seven steps, roughly four to eight weeks from first quote to working data, and one very expensive mistake you can avoid if you read step 3 carefully.

One thing before we start: this is a purchasing and coordination checklist, not an electrical one. I'm an office administrator, not a licensed electrician, and nothing here tells you how to wire anything. If a vendor proposes an install with no licensed electrician on site, thank them for their time and move on.

The 7-step checklist

Step 1: Start with the question, not the hardware

Most salespeople open with "how many meters do you need?" Don't answer yet. The number of meters is an output, not an input.

Write down what you're actually trying to answer, because the answers lead to different hardware:

• Cost allocation—which department or tenant used what. You need reasonably accurate per-feeder data, not lab-grade precision.
• Tenant billing—now accuracy matters legally and financially, and you'll want a platform with clean, exportable monthly reports.
• Savings baseline—you want to prove a retrofit or behavior change worked. You need a stable baseline period before you change anything.
• Equipment troubleshooting—odd spikes, tripped breakers, flickering lights. This is the one case where a power quality analyzer earns its price. If your problem is "the bill is too high," a power quality analyzer will not fix it, and anyone selling you one for that reason is guessing.

In my opinion, roughly half of the sub-metering projects that disappoint people start with the wrong question in step 1. Everything downstream just makes it worse.

Step 2: Walk the panels with your electrician

You need a licensed electrician for this walkthrough—not for the sales pitch, for the actual site survey. Your job is to document; theirs is to assess.

Things I bring: a phone for photos, a notebook, and a rough sketch of every electrical room. Things I ask the electrician to confirm, in writing:

• How many feeders need a current transformer (CT), and what size CTs fit the existing conductors
• Whether the panel has physical room for the meters and CTs—and whether the door will still close
• Whether any panel requires a shutdown, and if so, when that's allowed
• What permits and inspections the local authority requires

That last one surprises people. Permit requirements vary a lot by city and by building type, and a quote that excludes permits is not a cheaper quote—it's an incomplete one. Your electrician should know the local code requirements (the NEC in the US, or your local equivalent). I don't pretend to.

Also: find the panels nobody told you about. In our second building, the main panel was in a locked utility closet behind the mailroom, and the first vendor's quote missed it entirely. That closet cost us an extra visit and a change order.

Step 3: Ask "what's not included" before "how much"

This is the step that matters most, and it's the one almost every buyer skips.

If you ask me, transparency in pricing is worth more than a discount. A vendor who lists everything upfront—even if the total looks higher—usually costs you less in the end. I learned this the hard way, and here's the comparison that taught me.

Two quotes for the same 24-point project, both from companies that passed my basic checks:

• Quote A: $12,400. Looked great. It covered meters and CTs only.
• Quote B: $16,800. Covered meters, CTs, installation labor, after-hours work, commissioning, permits, and the first three years of platform subscription.

Quote A's real total after install labor, a gateway, permits, and a subscription came to about $19,000—and I spent three weeks chasing the missing pieces. Quote B was, honestly, the cheaper option. It took me about three years and two vendor changes to understand that the meter is the cheap part and the quote structure is where the money lives.

So the question to ask every vendor is simple: "Walk me through what's NOT included in this price." Then keep asking until the list stops growing. Watch specifically for:

• Installation labor and any after-hours or weekend premium
• CT wiring, conduit, and extension cable—often quoted per foot
• Gateways, network drops, and cellular data plans
• Platform subscription, and what happens to the price in year two or three
• Commissioning, verification, and training
• Permits, inspections, and any landlord approvals

Any vendor who gets defensive at that question is telling you something useful. Believe them.

Step 4: Price the data side—the part quotes hide

Nobody budgets for the software. Everybody should.

Sub-meter hardware in early 2025 ran me roughly $150 to $400 per point depending on CT type and whether it's single-phase or three-phase, before installation. The platform is a separate line item, usually billed annually, and it's the line item most likely to increase.

Ask, in writing:

• Is the subscription per meter, per site, or flat?
• What's the renewal price, and is it locked for more than one year?
• Does the platform work if you switch vendors later, or is the data locked in?
• Can you export raw data as CSV, and does the export include the timestamps you need?
• Where does the data live, and does your IT or security team need to review it first?

That last question has killed more projects than budgets have. Get IT involved in step 4, not step 6. Some energy monitoring IoT platforms—GridSense's power monitoring line is one example—bundle the gateway and dashboard into a single contract, which simplifies the accounting but means your data portability question matters even more. Ask it anyway.

And this was accurate as of early 2025. Pricing and platform terms move fast in this space, so verify current rates before you put numbers in a budget.

Step 5: Test the platform with your actual workflow

A beautiful dashboard is not a working system. Before you sign, get a demo or a trial and run three specific tests.

First, the report test. Ask the vendor to produce the exact monthly report you'd send to finance, using sample data. If they can't, or if it takes them a week, that tells you how your first month-end will go.

Second, the export test. Export a month of data to CSV and open it. Check the timestamp format, check whether interval data is included, check whether you can actually use it in a spreadsheet without cleanup.

Third, the access test. Set up a viewer account for a department manager and see what they see. If the permissions model is all-or-nothing, you'll be fielding "why can't I see my floor" emails for a year.

I'm not an IT security specialist, so I can't evaluate a platform's architecture. What I can tell you from a procurement seat is that the vendors who pass these three tests tend to be the ones who answer emails at month-end. That's the real product.

Step 6: Decide who sees the data before the meters go live

This is a people problem disguised as a technical one, and it's where projects quietly fail.

In one building, we published floor-level consumption to everyone. Within two weeks, departments were emailing each other about who was wasting power, and one manager asked me to "prove" his neighbor's usage. We didn't have the granularity to do it, and the whole thing damaged trust for months. Learn from that: decide the visibility model first.

Questions to settle with your operations lead and, if tenants are billed, your legal or finance team:

• Who sees site-level data, who sees floor-level, who sees nothing?
• Are tenants getting billed on this data, and does the lease say how?
• Do you publish a monthly summary, or is this internal only?
• What's the plan when someone disputes a number?

Get the answers in an email before installation. It takes twenty minutes and it prevents the worst meeting of your year.

Step 7: Put a review date on the calendar

Almost nobody does this step, and it's the difference between a system and an expensive ornament.

Ninety days after go-live, sit down with three questions: Are the numbers moving (i.e., is the data actually flowing)? Do the totals roughly match your utility invoices for the same period? And is anyone besides you logging in?

If the answer to any of those is no, you have a warranty or support conversation to have, and you want it while the vendor still cares. I have seen two installations where nobody checked, the gateway had been offline since week three, and the "project" was reported as complete for eleven months.

Pick the date now. Put it in your own calendar, not the vendor's.

Common mistakes worth avoiding

A short list, from my own errors and other people's:

• Buying a power quality analyzer for a billing problem. They solve different problems. Match the tool to the question in step 1.
• Comparing hardware prices instead of project totals. Quote A vs. Quote B in step 3 is the whole lesson.
• Ignoring the year-three subscription price. The hardware is a one-time cost. The platform is a relationship.
• Leaving finance out until the purchase order. They will ask the "what's not included" question eventually. Better in week one than month three.
• Letting the electrician pick the platform alone. Electricians are excellent at installation and understandably loyal to the interface they know. You're the one who has to run reports.

The bottom line: in a sub-metering project, the meters are the cheapest part. The quote structure, the data platform, and the review habit are where the money and the headaches actually live. If you take one thing from this checklist, take step 3—ask "what's not included" before you ask "how much." That question has saved me more budget than any negotiation.