Bear Method

Field diagnoses of small public companies, from the outside


Specimen No. 002Educational Development CorporationJuly 2026bearmethod.ai

The Company That Had to Give Its Name Back

Educational Development Corporation sold $204.6 million of children's books in fiscal 2021, almost all of it through a division named after a British publisher it had licensed since 1988. Then the publisher took the name back. Revenue is now $22.9 million, and Google's own suggestion box is still asking what the company is called. Here is the gap, measured.

What this is: a marketing and positioning diagnosis of a public company, built entirely from the outside: their SEC filings and open demand signals. We picked Educational Development Corporation because its last five years are the cleanest public demonstration we know of what borrowed brand equity costs when the owner wants it back, a risk most private companies carry without ever pricing it. What this is not: investment research. We say nothing here about the stock, and you should draw no conclusion about it from anything below.

The One Thing

In fiscal 2021, Educational Development Corporation booked $204.6 million in revenue. $196.0 million of it came through a division called Usborne Books & More: roughly 48,700 active consultants, on average, selling a British publisher's children's books at home parties and online party events under a license first signed in 1988. The name on every party invitation, every consultant's website, every cardboard box was Usborne. It was never theirs.

In May 2022 the publisher replaced thirty-four years of accumulated agreements with a narrower one, and the name went home. EDC lost the right to sell Usborne to retail stores, kept a multi-level-marketing channel, and rang the NASDAQ closing bell that December to launch the division's new name, PaperPie, an invented word with no search demand attached to it. Three fiscal years later, revenue is $22.9 million, down 88.8% from the peak, and the active sales force is 4,300, below the 6,000 that Craig White, then chief executive, said the division had "only 10 years ago," in the December 2022 announcement of the rename itself.

Here is the part that makes this a marketing story rather than a licensing story. Three and a half years after the rename, Americans still search the abandoned name at roughly twice the rate of the new one. The single most common thing they search alongside it is "usborne books and more," a division that no longer exists. The market is still asking for this company by its old name. Nobody is answering.

The Symptom

The numbers below are from EDC's own filings, fiscal years ending in February.

Table 1.Six years, total company
Fiscal yearFY21FY22FY23FY24FY25FY26
Net revenues, $M 204.6142.287.851.034.222.9
Net earnings (loss), $M 12.68.3(2.5)0.5(5.3)2.3
Active sellers at year-end 57,60036,10024,60015,0007,8004,300
Inventory, $M 51.871.659.143.929.117.4
Source: Forms 10-K FY2021–FY2026, SEC EDGAR, CIK 0000031667. Seller counts are each 10-K's approximate active consultants (through FY22) or Brand Partners (FY23 on) at fiscal year-end.

Read the bottom two rows together and you can watch a company eat its own pantry. Inventory bought for a 57,600-seller network has been sold down for four straight years to feed a 4,300-seller network, and the proceeds went to the bank. The two profitable years in that stretch were not operating recoveries: fiscal 2024's $0.5 million came from a $4.0 million gain on selling the old headquarters plus $3.8 million in pandemic-era employee retention credits, and fiscal 2026's $2.3 million came from a $12.2 million gain on selling the Hilti Complex, the Tulsa headquarters and distribution campus, for $29.9 million in proceeds in October 2025. That sale paid off all bank debt and lifted the going-concern doubt that had sat in the filings; it also means the buildings are gone. Since fiscal 2023, the profitable years were profitable because of what the company sold off, not what it sold.

The sharper symptom is the seller count, because in this business the seller is not a salesperson. She is the distribution, the storefront, and the advertising, all at once. EDC's total advertising expense in fiscal 2026 was $249,800, about $21,000 a month for the whole company, against $6.4 million paid in sales commissions. The marketing engine was always the network itself: 26 dollars of commission for every dollar of advertising. When the network went from 57,600 to 4,300, the company did not just lose sellers. It lost the only media channel it had.

And the license that started all of this is not finished moving. The remaining Usborne agreement carries annual minimum purchase volumes, and the fiscal 2026 10-K states it plainly:

During fiscal 2025 and fiscal 2026, the Company did not meet the minimum purchase volumes. No notification of non-compliance or termination has been received from Usborne. Educational Development Corporation, Form 10-K for fiscal 2026, filed May 19, 2026
Fig. 1.Active sellers at fiscal year-end Educational Development Corporation, FY2021 – FY2026. “Active” = sales in the past six months. Hover a bar for the year's events.
Source: EDC Forms 10-K, FY2021–FY2026, Item 1 (approximate year-end counts). SEC EDGAR, CIK 0000031667.
The Outside View

Four public signals, captured July 7, 2026.

The name they gave up still out-draws the name they own, two to one. On a shared Google Trends scale over five years, "usborne books" averaged 27.7 with a peak of 100 in November 2021; it sits at 10 now. "Paperpie" has averaged 10.0, peaked at 40, and sits at 5, with several 2026 weeks too low for Trends to report at all. The new name never rose. It appeared in early 2023 at about a third of the old name's already-fallen level and has declined alongside it since.

Fig. 2.The abandoned name vs the invented one, shared scale Google Trends weekly index, United States, July 2021 – July 2026. 100 = the busiest week for either term.
“usborne books”“paperpie”
View data table
Source: Google Trends multi-keyword comparison, captured 2026-07-07 via SerpAPI. Relative indices, not absolute volumes.

What people search next to the old brand is the old brand's fate. The top related query for "usborne books" over five years is "usborne books and more," the retired division name, at index 100. The fastest risers include "usborne books new name," up 1,700%, and "usborne books name change," up 700%. Buyers are not confused about wanting the books. They are confused about who sells them now, and they are asking Google instead of a Brand Partner.

The demand EDC built now lands on the publisher, Amazon, and HarperCollins. The "usborne books" search page has no ads and five organic results: usborne.com itself ("Shop direct with Usborne"), the publisher's Instagram, Amazon's Usborne brand store, HarperCollins's Usborne page, and a parenting blog. Educational Development Corporation and PaperPie appear nowhere on it. Three and a half decades of American living-room marketing built a demand stream that now flows entirely to other companies, including the brand's owner selling direct.

The new name is findable, unsought, and answering the door badly. PaperPie owns its own search page: paperpie.com ranks first for "paperpie." But the indexed snippet under that first result reads, verbatim, "An unhandled error has occurred. Reload." The related searches are "PaperPie back office," "PaperPie login," and "PaperPie reviews," consultant navigation rather than buyer demand, and Google's related questions still include "Is PaperPie the same as Usborne?" Three years in, the search engine is doing the brand transition work the company has not.

Two hypotheses the data killed. First, this is not a product-demand problem: Usborne product searches like "sound books" and "sticker books" are still rising, the Amazon brand store is active, and the publisher thought US demand worth taking back. The books did not stop selling; they stopped selling through EDC. Second, this is not just the COVID bubble unwinding: against pre-pandemic fiscal 2020, before the boom, revenue is down 80% and the average sales force is down 82%. The decline blew straight through the pre-bubble baseline and kept going.

The Diagnosis, Hypothesized

From the outside, this looks like a network that ran on a name, kept the network, lost the name, and then watched the network discover it could not run without it.

The mechanism has three turns. First, every one of those tens of thousands of Brand Partners was a micro-business whose single asset was recognition: "Usborne" was the word that got a party invitation accepted, a Facebook event joined, a school fair approved. The rename repriced that asset to zero for all of them on the same day. Second, in a party-plan business the recruiting engine is the network itself: hostesses become consultants, guests become hostesses. Fewer sellers means fewer parties, fewer parties mean fewer recruits, and the loop compounds. New Brand Partner additions fell from 7,800 in fiscal 2025 to 2,700 in fiscal 2026. Third, the survival plan fed the loop: lender restrictions kept EDC from reordering stock or introducing new titles for the first three quarters of fiscal 2026, so the sellers who stayed had less to sell. The company's own explanation for partner losses, in the fiscal 2026 10-K, names "our distribution agreement with Usborne whereby Usborne actively sells their products through discounted retailers in the U.S. market, and the rebranding of the division in the fourth quarter of fiscal year 2023."

What the mechanism did not break is telling. Revenue per average active seller was about $4,025 in fiscal 2021, dipped to about $2,427 in fiscal 2025, and recovered to about $3,335 in fiscal 2026: a band, not a collapse. The sellers who remain still sell at broadly historic rates. The product still works, the model still works at the level of one seller and one living room. The transmitter count collapsed; transmitter output did not. That is why we read this as a marketing diagnosis and not a business-model autopsy.

HibernatingEDUC · FY2026 · bearmethod.ai

Underperforming its market by an enormous margin while holding genuinely differentiated assets: the exclusive US MLM channel for Usborne, full ownership of Kane Miller, SmartLab Toys, and Learning Wrap-Ups, and a seller network that still converts. The differentiation exists; it is aimed through a name that carries no demand. Repoint the differentiation. Do not abandon it.

Management's stated plan (reorder key titles, introduce new ones, improve the Brand Partner back office) is operational and assumes the network re-energizes itself. The two forces that built the network, a name with ambient demand and a pandemic income shock, are both gone, and the filings do not name what replaces them. What we cannot see from outside, and say so plainly: the naming provisions of the 2022 distribution agreement, meaning whether and how PaperPie may publicly call itself the former Usborne Books & More; Brand Partner churn, cohort economics, and earnings distribution; the size and reachability of the customer file; average order values; the dollar terms of the Usborne minimum purchase volumes; and the school and book-fair channel's real economics. Any of these could revise the diagnosis. That is what inside data is for.

Three Repositioning Hypotheses

These are hypotheses, not recommendations. Each comes with the test that would confirm or kill it.

H1Say the old name, everywhere the agreement allows.

The search data says the most valuable sentence in this company's possession is "PaperPie is the new name of Usborne Books & More." People ask Google exactly this, 1,700% more than five years ago, and on the company's own brand search page the explainer that ranks is a Reddit thread. Put the sentence on the homepage, in title tags, in every Brand Partner's script, and in a small paid test against "usborne books" queries.

Test Eight weeks: measure branded-search transfer, party bookings, and reactivation of lapsed customers. Kill No movement in eight weeks, or the agreement's naming provisions bar nominative reference, which is the first thing we would read.

H2The book fair is the beachhead, not the party.

The party model needs a name the guest recognizes; a school book fair does not. Schools buy the catalog and the fundraising split, not the brand on the consultant's tote bag. EDC already has the program (PaperPie Learning) and a fair does at scale what parties did one living room at a time: it puts the product in front of parents and recruits the next seller.

Test A two-district pilot, twenty schools, measuring revenue per fair, the school's take, and Brand Partner recruits per fair against party-channel baselines. The pilot also answers whether schools actually want an alternative to Scholastic, which we hypothesize and cannot see from outside. Kill Fairs that produce neither margin after the school's cut nor recruits.

H3Treat the 4,300 as the asset, and retention as the marketing program.

In this structure every departing seller is lost distribution and lost advertising at once, the same double loss both this company and our first specimen keep teaching. The highest-leverage marketing spend is plausibly whatever makes the remaining 4,300 loudly successful, because sellers are recruited by the visible success of other sellers.

Test This one cannot be run from outside. It needs partner-level earnings, churn cohorts, and downline structure, which is exactly the conversation we would want to have.
Simulated · Synthetic Panel
What a Simulated Buyer Panel Said

Before publishing, we ran the positioning question past a synthetic buyer panel: ten constructed personas of children's-book buyers, from a mom with a shelf of party-bought Usborne books to a PTA book-fair organizer, an Amazon price-checker, an MLM skeptic, and two deliberate mismatches. This is simulated data. It is not customer research, and no EDC customer data exists in it. We use it the way an architect uses a wind model: cheap directional pressure-testing before anyone builds anything.

Fig. 3.Simulated conversion by message, weighted
“PaperPie is the new name of Usborne Books & More”Hypothesis H1, the heritage bridge
60.3%
“Kane Miller books and SmartLab toys, only from PaperPie”The owned-brands message
52.5%
“PaperPie: gather for good around literacy and learning”The current positioning
52.3%
Synthetic panel, 10 personas, weighted. The panel was deliberately seeded with buyers who remember Usborne, which favors the heritage message: trust the direction more than the size.

Unlike our first specimen, where the incumbent message held its ground, the incumbent lost here, by eight points to the sentence the company has stopped saying. The edge concentrated exactly where you would expect: the personas who remember Usborne. For buyers with no memory of the parties, the heritage line did nothing, which is the second finding: this equity is a wasting asset that only works on the cohort that remembers it, and that cohort ages out a little more every year the sentence goes unsaid. The owned-brands message, a plausible strategic fix, tied the incumbent on cold buyers, so the panel killed it as a message even if it stands as a long-term strategy. One caveat carries legal weight: whether PaperPie may say the winning sentence in paid media depends on naming provisions in the 2022 agreement that are not public. The panel tests pull, not permission.

The Panel.Ten buyers we invented so EDC would not have to guess
Book-Party Megan
34, Columbus OH, two kids, a shelf of party-bought Usborne books, does not recognize the name PaperPie
Skepticism low15%
Homeschool Rachel
38, rural Tennessee, three kids, builds curriculum from co-op recommendations, suspects MLM markup
Skepticism medium14%
Gift-Giving Linda
64, Cary NC, five grandkids, loved an Usborne dictionary from a church bazaar, wary of unfamiliar sites
Skepticism medium12%
PTA Organizer Jennifer
42, Fairfax County VA, ran an Usborne book fair in 2019, scouting a Scholastic alternative
Skepticism medium10%
First-Library Destiny
27, Atlanta, new mom building a nursery shelf, no Usborne memory at all
Skepticism medium9%
Price-Checker Marcus
36, Dallas suburbs, screenshots the ad and opens Amazon in the next tab
Skepticism high8%
Teacher Karen
45, Springfield MO, 2nd grade, spends her own money on classroom books, remembers a neighboring school's Usborne fair
Skepticism low8%
Anti-MLM Aisha
31, Portland OR, likes the books, refuses the business model on principle
Skepticism very high5%
Collector River
29, Seattle, collects illustrated books as art, has no children
Skepticism high3%
Gardener Harold
68, rural Alabama, no grandchildren, assumes unfamiliar Facebook ads are scams
Skepticism very high2%
Panels like this one are how we pressure-test a message before a dollar of media spends. They are built in SimPanel, our synthetic buyer panel tool: describe your customer, get a weighted panel, run your copy against it. See how a panel is built →
What We Would Verify With Inside Data

The naming provisions of the 2022 Usborne agreement, before anything else. Brand Partner churn by cohort and the earnings distribution across the 4,300 who remain. The customer file: how many buyers from the 57,600-seller era are reachable, and what a reactivation message returns. The dollar terms of the minimum purchase volumes and the realistic runway on the remaining license. Book-fair unit economics against Scholastic's split. Kane Miller and SmartLab contribution margins, since they are the assets nobody can call home.

That is a two-week diagnostic, not a two-year transformation.

This report is a specimen of the BEAR method: a marketing diagnosis we normally run for private companies, applied here to a public one because its filings let us show the work. If you run a company and want this run on your business, with your real data instead of the outside view, book a BEAR session. If you just want the next diagnosis when it publishes, subscribe.
Published by Click Makers, LLC. This is marketing and positioning analysis for educational purposes. It is not investment research, not investment advice, and not a recommendation regarding any security. We hold no position in EDUC, have no relationship with the company, and received no compensation related to this report. All figures come from the cited public sources as of the capture dates stated above; Google Trends values are relative indices, not absolute volumes. The synthetic panel results are simulated and are labeled as such wherever they appear. Usborne is a trademark of Usborne Publishing Limited. PaperPie, Kane Miller, SmartLab Toys, and Learning Wrap-Ups are trademarks or brands of Educational Development Corporation. If you are the company and believe any fact here is wrong, write to corrections@bearmethod.ai and we will review and correct promptly.
Bear Method · Specimen 002 · Set in the field-journal style · MMXXVI