In 1993, A.P.F. Turner published a surprisingly detailed history of glucose sensing's path from lab to market. I'm taking a break from designing our IIB course series to re-interpret this classic biosensing article from an economics and entrepreneurial perspective. Looking back, Turner identified several pervasive principles leading to success for early glucose sensors that are still quite relevant today on commercializing scientific innovations. Our IIB course series is built around combining these industry lessons with strategic management and entrepreneurial principles with just enough economics to make good choices in early-stage bioventures.

Turner's verdict is that glucose sensing (still a major portion of the overall biosensing market) succeeded in the 1990s because an urgent, chronic, evidence-backed demand met a chemistry and a production process cheap enough to serve it billions of times a year. The science was necessary but not the decisive scarcity.

The science side

The enabling idea was the mediated enzyme electrode. Immobilized glucose oxidase turned a plain platinum electrode into an analyzer; a small redox molecule (the mediator) then shuttled electrons from enzyme to electrode, freeing the reading from dependence on oxygen in the sample. Groups at Cranfield and Oxford put ferrocene to this use, and a three-person startup, Genetics International (later MediSense), brought it to market in 1987.

Turner's account of the chemistry is notably unromantic on two points.

• Patent luck. The ferrocene patent is, in his words, arguably the only one that has "truly stood the test of time," yet the original UK filing claimed chloranil. Ferrocene was added in a supplementary filing months later.
• The incumbent mediator won. Ferricyanide had been used in Roche laboratory analyzers since the late 1970s, and it became the mediator in most commercially successful strips. Boehringer Mannheim (later Roche) entered with a ferricyanide device bought from a small US firm, Tall Oaks.

Accuracy came from later engineering more than from the core reaction. Unilever's "Shanks" capillary-fill patent drew a pinprick sample into the strip and curbed evaporation and the cooling it causes, both of which bias readings. Subsequent gains, documented mainly in patents rather than journals, included automatic switch-on, temperature correction, fill detection and compensation for hematocrit (red-cell content), often inferred from viscosity.

Manufacturing and component sourcing

Turner ranks machine manufacturability as the second driver of success, after demand. Scaling is essential, especially for a physical product that needs mass production and distribution. Every commercial biosensor before MediSense's was assembled by hand, which is fine for custom durable-goods orders but cannot supply billions of disposables a year. The MediSense team adapted screen printing -- the technique behind T-shirts and circuit boards -- to lay electrodes and enzyme inks onto polyethylene terephthalate (PET) film. The team never patented the process for this use, though it later appeared within claims in many patents by MediSense and its rivals, a classic case of weak appropriability a la Teece (1986). Sic vos non vobis mellificatis, apes: MediSense made the honey, and the whole industry ate it.

The component story is one of commodities. A plastic sheet, printable carbon and metal inks, an industrial enzyme and an inexpensive mediator such as ferricyanide make a strip; a few dollars of electronics make a meter. Turner puts bulk meter cost at $7–10, rising to about $90 for elaborate models, and notes that makers effectively give meters away as loss leaders. He does not discuss enzyme supply contracts or supplier concentration; readers wanting that should consult Newman and Turner (2005).

His forward look extends the same logic: print the battery, display, antenna and much of the circuitry too, leaving only a speck of silicon. A Linköping–Acreo prototype already read glucose on a credit-card-like printed device.

Market uptake

Demand comes first in Turner's list, and rightly so. The Diabetes Control and Complications Trial showed that tight control, guided by frequent home testing, sharply cut blindness, kidney failure and nerve damage (Turner dates it to the mid-1990s; the results were published in September 1993). Testing became a daily obligation as their chronic disease turned patients into expert repeat buyers of the new technology. Because they tested several times a day for life, modest gains in convenience were noticed by consumers and paid for (frequently by third parties such as insurers, which reduced the incidence of major costs of hospitalization and long-term complications.)

Entry followed a recognizable pattern of small inventors and large distributors:

The resulting "big four" held about 90% of sales. Design mattered too: the ExacTech pen was so small that it required more dexterity and eyesight than users had. The pen, it turned out, was mightier than the patient's grip.

By 2013 Turner called the strip a mature product: chemistry had stalled and firms were harvesting earlier investment, while engineering moved to strip cassettes, 1–3 µL samples, gentler lancets and phone apps. Bayer's DIGIT even linked a meter to a Nintendo console, rewarding children with game levels for testing. Meanwhile the grail of non-invasive measurement stayed out of reach, Tantalus-fashion: of 96 firms in Turner's survey expecting to launch within two years, none did. Hard science does not necessarily lead to a soft landing.

Principles for success

Turner offers glucose as a template for other analytes. Read as an economist, it yields nine principles; the first five rest directly on his text, the rest are my inferences from it and match generalizable principles from the strategic management, entrepreneurship, and technology strategy disciplines (less technical than electrochemistry but, I would argue, equally important to value creation and capture in bioventures.)

1. Certify the demand. A trial showing that biosensor measurement changes health outcomes converted a nice-to-have technology into a universally-implemented standard of care. Before the DCCT, frequent testing was advice that was selectively followed; after it, a prescription item with much better compliance. Unsurprisingly, verifying demand is the very first step in lean-startup entrepreneurship itself (including our IIB offerings) from Ries (2011), even before the final design of the product is finalized. To participate in discussion on this topic, please join our IIB Forum Discussion on Demonstrating Demand.
2. Seek frequent, lifelong use by expert buyers. Repetition makes users discern, so incremental improvements earn revenue and sustain investments in a quality race. Third-party payment through insurance means that consumers' desire to adopt makes their ability to pay for upgrades less important as long as they produce clinical benefit. Acquiring and Defending Customers.
3. Prefer cheap, proven chemistry. Ferricyanide, a decades-old reagent, beat the elegant ferrocene in market share. "Good enough" at a lower cost dominates "best in class" at a higher one, even though journals yawn at such a process improvement.  Look for more on this theme in future columns (and IIB courses); it is common yet frequently ignored. To participate in discussion on this topic, please join our IIB Forum on Innovation Scope.
4. Treat manufacturing as a peer invention to science. Screen printing, not a new molecule, made billions of units possible. Process innovation determines the cost structure, which in turn drives pricing and competition. To participate in discussion on this topic, please join our IIB Forum on Process Innovation.
5. Price the reader at cost or below, price the consumable at a high margin. A $7–10 meter given away locks in a strip annuity, the familiar razor-and-blades model that made Gillette a multibillion-dollar business with much less science. There is sound economic logic for such a multipart tariff that, once explained, resonates in multiple markets. To participate in discussion on this topic, please join our IIB Forum on Pricing and Revenue Models.
6. Expect profits to go to complementary-asset holders. Abbott, Roche, J&J and Bayer collected most of the returns; Unilever earned royalties without ever selling a meter. This is Teece's (1986) prediction where imitation is easy and distribution is scarce; suppliers of the scarce complementary asset help themselves to the returns even if they made little of the initial investment, as in the MediSense production innovation discussed above. To participate in discussion on this topic, please join our IIB Forum on Strategy for BioInnovators.
7. Design for the user's hands and eyes. Miniaturization past the customer's dexterity loses sales. User interface matters most when consumers choose among rival technologies, especially when they do not pay extra for a superior experience. To participate in discussion on this topic, please join our IIB Forum on Design and Usability.
8. Fix the friction before chasing the moonshot. Capillary fill and smaller samples paid off; non-invasive glucose has absorbed fortunes and has still delivered no FDA approvals as of 2025 (Workman, 2025) although investments (and slow progress) are still being made 30 years after the basic innovation was brought to market. To participate in discussion on this topic, please join our IIB Forum on Value Creation and Capture.
9. Plan for maturity. Once the core chemistry is settled, competition shifts to integration, software and cost, and margins go to whoever does those best. These skills are not, generally, the same as those driving success in scientific innovation or even manufacturing engineering. To participate in discussion on this topic, please join our IIB Forum From Biostartup to Going BioConcern.

For a new biosensor, the screening question is how many of the nine it satisfies. Most frontier analytes fail principles 1 and 2 before chemistry is even tested.