Where each answer comes from, and what it is worth
1. Published equivalent
Ninety-one of the competitor part numbers here come off our own Automotive Product Cross-Reference chart (v.001, 4/2026), which maps AE 215, AE 230, AE 250, AE 260, AE 303 and AE 7010 against the seven collision-repair ranges we cross-reference. Twelve of those chart entries now also carry the maker's own datasheet values, which is what turned up the chemistry errors described further down. Those pairings are ours, not computed, and they outrank anything the scoring model produces. Five codes sit against more than one AE product because the same competitor number covers work-time variants; on those, pick by work time, not by strength. The chart carries no competitor spec values, so a published equivalent is an application claim, not a strength claim.
2. Spec match
For everything else, each AE grade is scored against the competitor's published datasheet. Chemistry is a hard gate: a two-part epoxy scored against a methacrylate keeps at most 45 percent of its similarity, against a polyurethane 35 percent. The rest is a weighted similarity over these dimensions, renormalised over whatever data actually exists:
Two further multipliers keep thin data honest. A score is scaled by how much of the model could be evaluated, and it is cut by more than half when the competitor publishes a lap shear and we do not. That is why AE Putty Stick, which has safety data sheets but no technical data sheet, never surfaces as a serious match.
3. Custom formula
When the best AE answer scores below 50 and there is no published equivalent, the tool stops guessing and prints a spec envelope instead: the chemistry, cure, mix ratio, form, open time, shear and peel targets and substrate list the product would have to hit. That is the brief you hand a formulator, and it is a more useful answer than a match that will fail on the bench.
The surface-preparation problem
This is the largest source of false comparison in the dataset, and it cuts against us. Our automotive sheets state a lap shear on aluminium at 77°F but do not state a surface preparation. Some makers report acid-etched aluminium, some use an FPL etch on their aerospace sheets, and others report a dry rag or IPA wipe with no abrasion at all. The same adhesive can read 25 to 40 percent apart across those preps — one competitor's own datasheet prints 3,200 psi etched and 1,850 psi abraded for the same product. Every match states which situation it is in. Adding a prep line to the AE datasheets would remove the ambiguity, and it currently costs us the argument.
Two tiers of strength data
Not every competitor figure in this set is equally solid. Most of the industrial grades come from datasheets that at least name the substrate the lap shear was measured on, even where they omit the preparation. Twenty of them come from a maker that publishes strength as a range with no substrate named at all. Those grades are here because their chemistry, mix ratio, viscosity, cure schedule and service range are all published and all genuinely useful for narrowing a search — but their strength number is a weaker claim than the rest of the dataset, and every one of their cards says so in plain words. Read a match against one of them as a chemistry and handling match that happens to carry a number, not as a strength comparison you could defend on a bench.
Where AE300 is different
AE300 is the one grade whose figures are quoted on as-received surfaces with no preparation at all — 3,450 psi aluminium, over 4,000 psi steel, substrate failure on fiberglass, PVC, SMC, ABS, PMMA and gelcoat. Against the minimal-prep structural acrylics on the market that is the strongest claim in the line, because it is measured the way the customer actually works.
What we do not publish
- No peel figure at all for AE 215, AE 230, AE 250 or AE 260, even though AE 260's own literature claims high peel. Competitors publish peel on most of these grades, so it is the first number a customer asks for after shear.
- No mixed viscosity for any of the automotive grades, only the two component viscosities.
- No performance data at temperature. Every AE figure is at 70–77°F. Several competitors publish a full temperature spread.
- AE 420's literature is internally inconsistent: it states a 4 minute working time and a 15–25 minute sand time, but labels its gel time row in seconds as 3–5. One of those is a typo and it is on a customer-facing sheet.
- AE Putty Stick has safety data sheets but no technical data sheet, so it cannot be matched on specification at all.
Why competitor names are not shown
Part numbers are what a customer quotes, so they are what the tool leads with. Each one carries a short maker code — enough to place the part without printing a competitor's name across the page. The full names are held in the search index, so typing a brand still finds its parts and what you typed stays in the box. This is a legibility aid, not a disguise: three of the makers are already initialisms, and a part number identifies its maker to anyone who knows the market anyway. The point is that we do not advertise for them, not that we pretend they are anonymous. Each result carries its application instead. Datasheet sources are held on file rather than linked, since the link itself names the maker.
Filtering the AE line
Search by spec ranks the AE line against a requirement: substrate, chemistry, mix ratio, viscosity, fixture time, cure time and container. Two of those are thinner than they look. Fixture time is published outright only for AE300 and two trim bonders; for the automotive grades the sand time is the nearest published figure and every card says so. Full cure at room temperature is published for the seven automotive grades only. A grade that cannot answer a filter is scored down rather than hidden, and the card lists what was missing, so a short list is never silently wrong. Container comes from the published pack sizes: cartridges cover eleven grades, AE300 is the only grade with a published pail and drum, and no datasheet mentions a tote.
The equine market, and why every hoof product returns a custom formula
Forty hoof-care products from three makers are in the set, and every single one resolves to an AE Custom Formula. That is not the scoring model failing — it is the model refusing to do something it would otherwise have done wrong.
Sixteen of the forty are two-part polyurethanes, and AE carries three of those. Left alone, the model would have returned a seventy-point match between a hoof packing material and an automotive urethane on the grounds that the chemistry family agrees and the open times are in the same range. That match would be wrong in a way no number on the card could convey. An adhesive going onto a hoof bonds to living keratin on the weight-bearing limb of a half-tonne animal, has to move with a hoof capsule that flexes on every stride, and must not peak hot enough to damage the laminae underneath. No AE grade has ever been tested against any of that, so the tool applies a qualification gate and says so, rather than producing a score that implies otherwise.
The data here is also unlike the rest of the set. Not one maker in this market publishes a lap shear or a peel figure, because a farrier does not buy on those numbers. What they publish is set time, durometer, pack size, and — on six of the forty — a peak cure temperature. That single exotherm figure is the most useful number in the whole equine dataset and the only one with a welfare consequence, which is why it gets its own line on every card whether the maker published a number or only a claim.
Two things follow for AE. The first is that these forty cards are a formulation brief rather than a sales tool: the envelope on each one states what a hoof product would have to do, in the terms this market actually specifies. The second is that every maker here dispenses from a dual cartridge and a static mixer, which is a MixTips conversation available now, independent of whether AE ever formulates anything for a hoof.
An error we found in our own cross-reference chart
Harvesting one competitor's datasheets turned up three rows where our printed chart and the maker's own sheet disagree about what their product is.
- 8409 and 8440 — our chart records both as two-part polyurethane and pairs them against AE 303, which is a polyurethane. The maker's own datasheets call 8409 “an extra fast set, two-part epoxy adhesive filler and surfacer” and 8440 “a fast-set epoxy repair compound”. Both are epoxies. As printed, the chart offers a polyurethane as the equivalent of an epoxy.
- 8422 — our chart says two-part epoxy, the datasheet says urethane-modified epoxy. That one is a refinement rather than a contradiction.
The likely cause is visible in 8409's own datasheet, which says the product is “designed for bonding flexible/semi-flexible substrates such as polyurethane”. Read quickly, the word polyurethane comes out of the substrate line and lands in the chemistry column. If that happened on two rows it can have happened on others, so the chemistry column is worth re-checking across the whole chart rather than correcting two cells.
The pairing itself has been left alone. It is AE's own commercial judgement and 8409 may still be the right product to sell against AE 303 — but the chemistry shown now comes from the maker's datasheet, and both affected cards say what the chart claimed and what the sheet says.
Known capability gaps
- No one-part heat-cure epoxy. The competitor grades in that class — 2214, E-214HP, AV 118 — resolve to an AE custom formula.
- Nothing claims untreated polyolefins the way DP8005, DP8010 and TA4610 do. AE 420 bonds PP and TPO but not on the same terms.
- Nothing in the line is a potting or encapsulation compound.
- AE 260 is the strongest grade on paper at 4,240 psi, but its own datasheet excludes structural components — no core supports, rails, pillars or rocker panels. The tool flags that on every AE 260 match rather than letting the number speak alone.
Competitor figures on the industrial side were read from the manufacturer datasheets linked on each card; eight were independently re-checked and three needed correction. Nothing here substitutes for a bonded coupon on the customer's actual substrate and prep — it gets you to the right one or two products to sample.
