This post isn't a quarterly data refresh — the scoring window is still Q3 2024 through Q2 2026, the same period our July 22 update covered. What changed is how we compute two of the three pillars behind that same window. Playing by the Rules now judges a company against its own enforcement record instead of ranking it against everyone else's, so a company's score reflects what it actually did, not how it happened to compare to the rest of the index this quarter. And we've substantially expanded our research into what private-equity ownership costs a company — leveraged-buyout debt, dividend payouts to sponsors, and asset sales used to fund those payouts rather than the business — though, like all of our private-company data, that coverage is still limited by what we can find a real source for. We also tightened how stock buybacks are calculated, refreshed revenue figures across the index, and added 21 new companies.

Of the 162 companies on the site before and after this update, 87 scores went up and 56 went down; 19 were essentially unchanged. The median move was about 5 points, and 29 companies changed tier. That brings the total we track to 183. Higher scores are better, as always. We also shipped several changes to the site itself this cycle — more on those near the end of this post.

Until this cycle, Playing by the Rules ranked companies against each other, so a company's score depended partly on how everyone else in the index was doing — not just on its own record. That produced two odd outcomes: a company with a genuinely clean enforcement record could still score below 100 for no reason other than sharing the index with other clean companies, and a company with real, serious violations could look "fine" if the rest of the index looked worse. Neither of those outcomes says anything true about what the company actually did.

We fixed that. A company is now judged against its own record on an absolute scale: a confirmed zero-penalty record scores 100, full stop, regardless of what anyone else did this quarter — and heavier penalties, relative to the company's size, pull the score down by degree.

This is the single largest driver of upward movement in this update. eBay, Nike, Razer, Tapestry, Sprouts Farmers Market, Ulta Beauty, TikTok, and Staples each reached a perfect 100 with confirmed-zero penalty records, while Sony, Lululemon, and Menards rose sharply into the 90s under the same change.

Separately, we made the 24-month scoring window itself more precise — checking each penalty's exact date instead of approximating by calendar year. That's why a number of companies' counted violation totals shrank even though nothing was removed from the record: older penalties that only looked in-window under the old calendar-year approximation are now correctly excluded. Home Depot's in-window total fell from $98.8 million to $22.2 million and Walgreens' from $633.7 million to $514.4 million as a result, among others.

Scoring Wealth Extraction for a private company is harder than for a public one — there's no quarterly earnings call, no SEC-mandated pay-ratio disclosure, no public buyback report to check. This cycle we built out a much deeper way to capture what actually happens inside a private company's ownership: leveraged-buyout debt loaded onto the company by its owners, dividend recapitalizations — where the company borrows money specifically to fund a payout to its owner, the private-market equivalent of a debt-funded stock buyback — and asset sales used to fund payouts rather than the business itself. An ordinary dividend paid out of profits, with no new debt involved, isn't what we're describing here or scoring against; we treat public companies the same way, penalizing buybacks, not dividends. This brings private-company Wealth Extraction much closer to how thoroughly we already cover public companies — though it's still bounded by what we can find a real, checkable source for. Every event we do score ties back to one — a court filing, an SEC document, a settlement — and we label how confident we are in each one on the company's page.

The companies with the largest Wealth Extraction drops this cycle are almost all private companies where this deeper research now applies: Michaels, Sears, PetSmart, Sweetwater, Bass Pro Shops, Hobby Lobby, New Balance, and Hudson's Bay are among the most visible — see the specific stories behind Michaels, Sears, PetSmart, and Sweetwater below.

We also improved how we estimate CEO pay ratio for private companies with no disclosed number of their own. Previously every company in a sector got the same flat estimate; the estimate is now scaled to each company's own revenue relative to similar companies that do disclose. That change affects 53 companies this cycle.

Companies sometimes report stock buybacks as a running multi-year total rather than what happened in a specific period, which can make a buyback look bigger or more recent than it was. We now hold every buyback figure to a strict 24-month window and reject those running totals. GameStop, Macy's, and Big Lots are among the companies with visible positive moves once their buyback activity was measured correctly.

Revenue figures were updated for a number of companies this cycle. Several companies received sourced revenue figures where confirmed data had not previously been available in our records — including Zara, Mercari, and ASOS — with each figure cited on the company's score card. Previously-cited estimates were also revised for New Balance, Hy-Vee, Neiman Marcus, Nintendo of America, and others, with sources noted. Where a previously cited figure could not be confirmed against our sourcing standard, we apply the same principle as prior cycles: no number is better than a bad number, and those companies score neutral on revenue-dependent factors with an insufficient-data label on their pages.

Sam's Club, Haier Smart Home, GE Appliances, Skechers, AT&T, Audible, LG, Birkenstock, bookshop.org, OnePlus, Mint Mobile, Consumer Cellular, Libro.fm, Uncommon Goods, Bonanza, Lego, Dick Blick, Fender, Jerry's Artarama, Reverb, and Cricut now have fully scored profiles, bringing the total number of companies we track to 183.

Scores below are rounded and measured from the July 22, 2026 data push. Tiers: Trusted 75+, Moderate 50–74, Concerning 25–49, Avoid below 25.

Beyond the score changes above, we shipped several changes to the site itself. Category pages let you browse and compare companies within the same market — pet supplies, sporting goods, grocery, and more — instead of only ever looking up one company at a time. Every retailer page now includes a Better Alternatives section, surfacing higher-scoring companies in the same category so switching is one click away. Company pages also now show a score history chart, so you can see how a score has moved over time instead of only where it stands today. And the full company directory now supports instant search — results filter as you type.

All scores reflect a 24-month window ending in 2026 Q2. The next quarterly lobbying-disclosure deadline falls in October 2026; once those Q3 filings are published and ingested, the next update will incorporate fresh data and get its own entry here.