The price of trust: why verified reviews and identity are economic infrastructure

A tax nobody voted for

Every market runs on claims. A shop says it is open until nine. A seller says the phone is new. A clinic says it is licensed, a caller says she is from the bank, and a page on a map says the business at this address is the one you were looking for. Most of these claims are true. The trouble is that a buyer often has no cheap way to tell which ones are not.

When checking is expensive, people stop trusting and start compensating. They insist on cash on delivery, travel to see a product in person, ask a cousin who knows someone, or buy only from the largest names. Each of those habits is rational, and each is a cost. Together they work like a tax on every transaction, paid by honest businesses as much as by their customers, and collected by no one.

This article sets out what the public record says about fraud and impersonation in Ghana and its neighbours, why the reported losses understate the damage, how other markets have started to treat fake reviews and fake identities as matters of law rather than etiquette, and what organisations can do now. The argument is simple: verifiable reviews and verified identity are not product features. They are economic infrastructure, in the way that roads, title deeds and payment rails are.

What the records show

Ghana's Cyber Security Authority (CSA) publishes the most useful numbers. In the first half of 2025 it recorded 2,008 cyber incidents, up 52% on the same period a year earlier, with GH₵14.94 million lost; online fraud and impersonation accounted for more than 90% of the losses.¹ The pace has not slowed. Between January and July 2026 the Authority recorded 3,876 incidents, and online fraud alone made up 47% of them.²

The banking system shows the same pattern. The Bank of Ghana's fraud report for 2025 counted 24,778 cases, 48% more than in 2024, with GH₵101 million at risk.² In October 2026 the central bank issued a twenty-point public advisory that warned against impersonation accounts on social media, fake websites and fraudulent links, and reminded the public that no legitimate bank, mobile money operator or payment provider will ask for a PIN, password or one-time code by call, text, email or social media.³

Ghana is not an outlier. In South Africa, the banking industry's risk centre reported that digital banking fraud incidents roughly doubled in 2024, from 31,612 to about 64,000, with losses rising from R1 billion to more than R1.4 billion. It attributed them to social engineering that exploited human error, not to technical breaches of bank systems.⁴ In Nigeria, the interbank settlement system reported fraud losses at financial institutions of ₦25.85 billion across 67,518 cases in 2025; losses halved from 2024, when a single incident distorted the total, but social engineering remained the dominant technique.⁵

Across the continent, INTERPOL's 2025 Africa Cyberthreat Assessment found online scams, particularly phishing, to be the most reported cybercrime, with suspected scam notifications up by as much as 3,000% in some countries.⁶ Its Operation Serengeti 2.0, run between June and August 2025, led to 1,209 arrests and identified nearly 88,000 victims.⁷

The common thread is not clever code. It is people pretending to be someone they are not, and buyers with no quick way to check.

The fake listing problem

One category deserves particular attention because it turns a trusted tool against its users. In April 2026 the CSA issued an alert about fake business listings on search engines and maps. Between 1 January and 26 April 2026 it recorded 54 cases in which criminals had planted or altered listings, so that people searching for a real business called a fraudster's number instead, with GH₵266,195 lost.⁸

The sum is modest beside bank fraud. The mechanism is not. A customer who searches for a bank branch, a courier, a hospital or a hotel and calls the number that appears has done exactly what a careful person should do. The fraud succeeds because the listing borrows the authority of the platform and the reputation of the real business at the same time. The victim loses money; the business loses a customer and some of its name.

Online shopping fraud follows the same logic. In the first ten months of 2025 the CSA recorded 266 cases and more than GH₵600,000 in losses, through social media pages and messaging channels that presented themselves as shops.⁹ The goods were often real somewhere. The seller was not.

Both kinds of fraud exploit the gap between a claim and the means to check it. A listing claims to be a business. A page claims to be a shop. A review claims to come from a customer. Where nothing ties the claim to a verified identity or a verified event, the cheapest claim wins.

Why reported losses understate the cost

Reported fraud losses are a floor, not an estimate, for three reasons.

Most fraud is never reported. INTERPOL found that only 30% of the African countries it surveyed have a system for reporting incidents, and 95% said they lacked adequate training, resources and tools.⁶ In the mobile money industry, more than 70% of providers told the GSMA they consider law enforcement ineffective against fraud, citing a lack of technical capacity.¹⁰ Victims who doubt anything will happen tend not to file.

The larger cost is the trade that never happens. A buyer who has been burned once, or who knows someone who has, adds a margin of caution to every later purchase. They pay more to buy from a known name, spend time and transport to inspect goods, or simply do without. None of that appears in a fraud report, yet it is where most of the economic damage sits: in transactions that are slower, smaller or abandoned.

Low trust favours incumbents. When buyers cannot check a newcomer's claims, they default to the brands they already know. That protects the large and penalises the small, which is the opposite of what a growing economy needs. A young business with excellent service and no way to prove it competes on price alone.

The flip side is that trust, once it can be checked, is valuable. In a February 2026 survey of 1,002 adults in the United States, 97% said they read reviews when choosing local businesses, 41% said they always do, up from 29% a year earlier, and 93% said they had bought something after reading reviews.¹¹ No equivalent survey exists for Ghana, and figures from one market should not be assumed for another. But the direction is plain: where reviews can be trusted, people use them to decide, and that shifts business towards whoever serves customers best.

How other markets are pricing fakes

For years, fake reviews were treated as a nuisance for platforms to manage quietly. That has changed.

In August 2024 the United States Federal Trade Commission adopted a final rule, by a 5–0 vote, banning fake reviews and testimonials, including those generated by AI; buying positive or negative reviews; undisclosed insider reviews; company-controlled sites posing as independent; and the suppression of reviews through threats.¹² In December 2025 the Commission sent warning letters to ten companies about possible breaches, noting that civil penalties can reach $53,088 per violation.¹³

The United Kingdom went further in its consumer law. Its competition regulator, the CMA, secured undertakings from Google in January 2025 to detect and remove fake reviews, to place warnings on UK businesses that inflate their ratings and to ban repeat offenders, noting that as much as £23 billion of UK consumer spending a year is potentially influenced by online reviews.¹⁴ Since April 2025 the regulator has been able to fine firms directly, and in March 2026 it opened investigations into five businesses over practices such as leaving one-star reviews out of ratings and rewarding customers for five-star reviews without saying so.¹⁵

The platforms now report enforcement at industrial scale. Google says it blocked or removed more than 292 million policy-violating reviews and more than 13 million fake business profiles from its maps in 2025.¹⁶ Trustpilot says it detected and removed 7.8 million fake reviews in the same year.¹⁷

Two lessons follow. First, fakery is not a local quirk: it appears wherever reviews and listings have value, which is to say everywhere. Second, the world's response has converged on the same principle. A review is worth something only if a reader can tell how it was checked, and the people who game the system should face consequences that outweigh the gains.

Trust as infrastructure

Infrastructure is what lets strangers deal with each other without each one rebuilding the means from scratch. A land registry means a buyer does not need to know the seller's family. A payment rail means a merchant does not need to count cash. Verified identity and verifiable reviews do the same job for claims.

Ghana already has part of the foundation. The National Identification Authority reports that, as at 9 October 2026, it had enrolled 20,285,071 people and issued 19,291,527 Ghana Cards.¹⁸ The card already anchors SIM registration and financial accounts. What is missing is the everyday layer that lets a customer, a supplier or an employer check a claim in seconds: that this page belongs to that business, that this person works there, that this review came from someone who was actually served.

Three design principles separate trust infrastructure from trust theatre:

• Every claim says how it was checked. A badge that says “verified” without saying what was verified is a decoration. A review that says it followed a recorded visit, a confirmed purchase or nothing at all lets the reader weigh it.
• Money never buys the outcome. Where payment can change a score, a ranking or the order of reviews, readers learn to discount all of it. Paid tools for businesses are fine; paid outcomes are not.
• Identity is checked once and shown sparingly. Verifying an identity does not require publishing it. The useful signal is “this account is a real, checked person or business”, not the person's ID number.

What organisations should do now

None of this requires waiting for new law. Any organisation that deals with the public can lower the cost of trusting it, starting this quarter.

1. Claim and check your listings. Search for your organisation on the main search engines and map services every month. Claim each listing, correct the phone numbers and addresses, and report impostors. The CSA's alert shows that criminals rely on businesses not looking.⁸
2. Publish one place to verify you. A single page on your own website that lists your official phone numbers, social accounts, payment accounts and branch addresses gives customers something to check against. Link to it from every listing.
3. Say what you will never ask for. Tell customers, in plain words, that your staff will never ask for a PIN, password or one-time code, echoing the central bank's own guidance.³ Put it on receipts, posters and your website.
4. Collect reviews only from real customers, and never buy them. Invite reviews after a verified visit or purchase, not from friends or staff. Do not offer discounts for positive ratings, and do not hide negative ones. The rules now enforced in the United States and the United Kingdom show where expectations are going, wherever you trade.¹²¹⁵
5. Answer reviews in public. A calm, specific reply to a fair complaint is evidence of good faith that no marketing can buy.
6. Verify the people you pay. Before changing a supplier's bank or mobile money details, confirm by calling a number you already hold, not one in the message asking for the change.
7. Train the front line. Social engineering beats systems by persuading people.⁴ A short briefing on impersonation, repeated each quarter, is the cheapest control most organisations will ever buy.
8. Report what you see. Report fraud and fake listings to the Cyber Security Authority and the platform concerned. Reports are how the public record, and the case for better tools, gets built.

The organisations that make themselves easy to verify will win trade from those that do not. Over time, so will the markets that make verification cheap for everyone.