Billing disputes drove the Q1 2026 complaint mix. That is the first thing worth knowing before any electric company name comes up. The Public Utility Commission of Texas (PUCT) sorts every consumer complaint into a handful of categories, and in the January through March window, billing and payment issues were the single largest bucket. Service quality and disconnection complaints trailed behind. Switching and slamming complaints, the category that dominated headlines a decade ago, kept shrinking.
This article does two things. It reports what the latest PUCT snapshot says about complaint volume and category. Then it shows the math that separates a company that gets a lot of complaints from a company that gets a lot of complaints per customer, because those are different problems and most coverage conflates them.
The headline number, and why it needs context
PUCT publishes complaint data in quarterly snapshots, and Q1 2026 is the latest one available as of this writing. The Commission’s Consumer Protection Division processed several thousand retail electric complaints in the quarter. That sounds alarming until you set it against the size of the market. Texas has roughly 9 million retail electric accounts in the deregulated area. A few thousand formal complaints against 9 million accounts is a complaint rate well under one tenth of one percent.
That does not mean the market is healthy. It means formal PUCT complaints are the tip of a much larger iceberg. Most billing disputes get resolved (or abandoned) at the company’s own call center long before anyone files with the state. A formal complaint is what is left after the customer tried the company directly and got nowhere. So the PUCT count is better read as a measure of how often companies fail to fix problems themselves than as total customer dissatisfaction.
Keep that framing. A rising formal complaint count can mean more problems, or it can mean the same problems handled worse at the first line of defense.
Billing is the category that matters this quarter
Within the Q1 2026 mix, billing and payment complaints led. The specific sub-issues show up year after year: estimated meter reads that overshot actual usage, promotional rates that expired into much higher month-to-month rates without a clear heads-up, and bill-credit plans where the customer missed the usage threshold by a few kilowatt-hours and lost a credit worth more than the power they used.
That last one deserves a number. A common bill-credit plan structure gives a 100 dollar credit at 1,000 kWh of monthly usage. A household that lands at 999 kWh gets nothing and pays full rate on every one of those kilowatt-hours. The cliff is real, and it generates complaints because the customer feels tricked by their own thermostat. LightCompanies has flagged this plan design before. It rates poorly on rate transparency even when the company behind it has a clean service record.
The takeaway for a shopper: billing complaints are mostly a plan-design and disclosure problem, not a keeping-the-lights-on problem. A company can run a reliable grid relationship and still generate a wall of billing complaints because its plan menu is built on fine print.
The math that changes the ranking
Here is where most complaint coverage goes wrong. It publishes a list of the companies with the most complaints, and the list is almost always just a list of the biggest companies. Of course the company with 2 million customers logs more complaints than the company with 80,000. More customers, more complaints. That tells you nothing about quality.
The fix is to normalize. Divide complaints by customer base and express the result as complaints per 100,000 customers, or per 1,000 if you prefer a smaller denominator. Now you are comparing rates, not totals, and the ranking can flip hard.
Work a simplified example with round numbers to see the mechanism. Say Company A logs 600 complaints in the quarter and serves 2,000,000 customers. Company B logs 300 complaints and serves 250,000 customers. By raw count, A looks twice as bad. Normalize:
- Company A: 600 / 2,000,000 = 30 complaints per 100,000 customers.
- Company B: 300 / 250,000 = 120 complaints per 100,000 customers.
Company B is four times worse per customer, even though it filed half as many complaints in absolute terms. The raw-count list had them backwards. This is the single most important habit when reading any complaint data: ask for the denominator. If a ranking does not divide by customer base, it is mostly a size chart wearing a quality costume.
PUCT does not publish a clean per-customer rate for every company, and exact customer counts by company are not all public, so any normalized ranking carries estimation error. That caveat is real and it is disclosed here on purpose. The direction of the correction, though, is reliable: big incumbents look worse on raw counts and better on normalized rates, and small aggressive-marketing companies often look fine on raw counts and worse once you divide.
How the most-complained-about REPs actually sort
Apply the normalization lens to the usual comparison set and a pattern holds across quarters. The large legacy retailers, Reliant and TXU among them, carry high raw complaint totals because they carry millions of accounts. On a per-100,000-customer basis their rates tend to land mid-pack, not at the bottom. Their billing infrastructure is mature and their complaint-resolution processes, while not loved, are staffed.
The companies that rank worst on normalized billing complaints are more often the mid-size and smaller retailers competing hardest on teaser rates. The plan designs that win the click on a rate-comparison page (the steep bill-credit cliffs, the low advertised rate that assumes one specific usage level) are the same designs that generate billing complaints when real households use real amounts of power. The marketing and the complaint rate come from the same root.
LightCompanies does not publish a single most complained about REP label, because the answer depends entirely on whether you mean raw count or normalized rate, and the two produce different names. Anyone who hands you one name without telling you which measure they used is selling something. The honest answer is two lists: a raw-count list dominated by the biggest companies, and a normalized list where smaller teaser-rate retailers move up.
Reading it through the five-factor lens
LightCompanies scores every company on the same five factors. Q1 2026 complaint data feeds three of them directly.
Rate transparency. This is where billing complaints land hardest. A company whose complaints cluster around expired promo rates and bill-credit cliffs scores down here regardless of how polite its call center is. The complaint category is the evidence.
Billing reliability. Estimated-read disputes and double-billing complaints hit this factor. These are execution failures, not disclosure failures, and they separate companies with sound meter-data processes from companies still cleaning up after a billing-system migration.
Customer service responsiveness. Remember the iceberg. A formal PUCT complaint is what survives a failed call-center interaction. A high formal complaint rate relative to peers is itself a signal that first-line service is not resolving issues. This factor reads the gap between total dissatisfaction (unknowable) and formal escalation (counted).
The other two factors, plan flexibility and renewable mix, are not visible in complaint data and get scored from plan menus and fuel-mix disclosures instead. Complaint data is one input to the score, not the whole score. A company can post a clean complaint quarter and still rank low on flexibility if every plan locks you into 24 months with a steep early-termination fee.
What to do with this if you are shopping or stuck
If you are choosing a company, do not let a scary raw complaint headline drive the decision, and do not let a low one reassure you either. Ask three questions. What is the complaint rate per customer, not the total? Which category dominates that company’s complaints, billing or service? And does the plan you are about to sign carry the exact design (the credit cliff, the expiring promo) that generates those billing complaints in the first place? The third question is the one you control, because you pick the plan.
If you are already with a company and frustrated, the data points to a sequence. File with the company first and get a ticket number, because PUCT will ask whether you did. If that fails, file the formal PUCT complaint. The Commission’s own quarterly counts exist because people escalate, and escalation is what moves a company’s normalized rate in the next snapshot. The complaint you file is a data point the next shopper reads.
Where this data is thin
Three limits are worth stating plainly. PUCT publishes quarterly, so Q1 2026 is a three-month window and a single quarter can swing on one billing-system problem at one company. Customer-count figures by company are not uniformly public, so normalized rates are estimates with real error bars, not precise rankings. And formal complaints undercount total dissatisfaction by an unknown but large multiple, because most disputes never reach the state.
Nnone of that makes the data useless. It makes it a starting point that has to be read with its denominator attached and its category broken out. Raw totals track company size. Normalized rates track company quality. Billing led the Q1 2026 mix, which means the smartest defense a shopper has is not picking the company with the fewest complaints. It is reading the plan fine print that produces most of those complaints in the first place.