Every fix the industry sells now is a better audit, and MIT Sloan just published the best one I have read in years — which is exactly how I know the audit is not the architecture.
In October, MIT Sloan Management Review ran a piece asking whether HR technology is undermining corporate strategy. Two researchers at USC’s Marshall Center for Effective Organizations walked the entire problem: companies accumulate HR platforms the way restaurants accumulate dead rules, each one added to fix something local, none of them designed together. Vendor systems are fit for sale, not fit for purpose. Managers satisfice — they accept close-enough defaults until the defaults become the operation. Entire roles now exist as human glue, people hired to manually extract, translate, reconcile, and transfer data between systems that were never designed to talk to each other. They are right about all of it. They are describing the [Summers Principle] in a suit: outcomes are either designed or they are defaulted. Then they hand you the fix, and the fix is where the piece proves my point instead of theirs.
What The Article Gets Right
The evidence the researchers gather is the strongest part of the piece. SHRM found that fewer than half of HR professionals and employees regard their company’s HR technology as effective — with more than 400 vendors showing their wares at the last HR Technology Conference. The article names the condition correctly: the issue is not a lack of technology but an over-accumulation of platforms. Few stacks emerge from intentional architectural design. Systems evolve through the steady addition of point solutions, each one defensible in isolation. That is the default half of the [Summers Principle] stated in full, in an academic journal, by two researchers who have never read my work. When outside thinkers arrive at your mechanism independently, that is not flattery. That is confirmation the mechanism is real and spreading. In 45 years I have watched it arrive under a dozen names, and it is always the same principle wearing new clothes.
Good Is Not The Standard
The article’s word is satisficing — accepting close-enough defaults, categories, thresholds, workflows, and definitions. My version is shorter: good is not the standard. A close-enough tool meets the expense and speed requirements and quietly sets the ceiling on everything built on top of it. The researchers even name the compounding: each small compromise embeds itself in everyday practice, and a patchwork of suboptimized processes skews what the organization prioritizes. That is [Static Decline] in a systems vocabulary — their own image is wearing shoes you have outgrown for the rest of your life because they fit you once. The workflow still runs. The data is still generated. The logic quietly sabotages the outcome. Motion without achievement is the defining cost of the defaulted operation, and no industry has a monopoly on it.
The Check-In That Stops One Step Short
Then comes the fix, and this is the part worth the price of the piece. The authors offer a four-step diagnostic they call the Strategy-First Check-In: a blank-page exercise where you design the ideal system with no close-enough constraints; an ecosystem inventory where you list every tool; an accumulation audit where you map each tool’s data and decision logic; and a strategy test where you check the whole stack against your priorities. All four steps inventory. Not one step assigns. There is no fifth step that says: write down the name of the person who owns each decision the system surfaces, and who closes the loop when the number moves. The check-in ends at alignment. A strategy-aligned stack with no owner is not a fix. It is a cleaner dashboard.
Why The Reading Stops There
The stop is not a failure of reading. The researchers did the hard work — they got the vendor defaults, the satisficing, the human glue, the static logic. They walked all the way to the door. What they cannot do through the page is the last step, because the last step is not a reading problem. The audit is the layer everyone can sell: questionnaires, frameworks, dashboards, alignment maps. The act layer — who owns the decision — is a decision, and decisions are operator territory. The piece even frames defaulting as an information problem: the trade-offs are not immediately obvious, so surface them with better questions. Defaulting is not an information problem. It is an ownership absence. No questionnaire fills it, because a question cannot sign its name.
The Same Close Everywhere
This is the same close everywhere, and it is why I teach the principle instead of the audit. Menu engineering computes the ROI of the line and then ignores it. Industry reports get read for the data when the sponsor list is the design. Ordering platforms promise you all your data and share the archive, not the flow. In every case the diagnostic is excellent and the loop is unassigned. The questions a tool asks are shaped by who built it — [Vendor Capture] — and the one question no vendor’s tool asks is the one that ends the capture: whose name goes on the decision?
The Name On The Loop
What a designed operation does that an audited one cannot is put a name on the loop. The standard is carried by the design, and the design includes a name — the Host who can fix the visit slipping, the kitchen manager who can flex the line, the role closest to the work holding the authority to act. When the truth only surfaces when the owner walks in, the design is hiding it from the people who work it every night. An audit can tell you that is happening. Only the design can stop it, because stopping it is a decision, and decisions have names on them.
The Owner Test
Run the owner test on your own operation. For every report, dashboard, and audit you currently receive, answer three questions. Name the person who acts on it. Name the decision it drives. Name what changes Monday when the number moves. Any blank line is not a reporting gap. It is a default wearing a diagnostic’s clothes.
The Only Field That Matters
The audit tells you where the design is missing. Only the design tells you whose job it is to close the gap. The field marked Owner is the one no vendor defaults, no platform inventories, and no check-in surfaces — and it is the only field that matters. The motion continues either way. Whether it belongs to you is decided by one thing: whether a name is on it.
What You Do Monday Morning
Take the one report you actually read every week. At the top of the page, write the name of the person who acts on it. If you cannot write a name, you do not have a reporting problem. You have found a default — and now you know exactly where the design work starts.
Digging Deeper
- What I Sell — jeffreysummers.com
- How I Use AI — jeffreysummers.com
- Why Good Employees Aren’t Ever Enough — Restaurant Physics
- The Only Hour That Multiplies — Restaurant Physics
Terms used in this piece: Summers Principle, Static Decline, Vendor Capture. Definitions in the Knowledge Base.
Sources Cited In This Piece
- MIT Sloan Management Review, October 26, 2026 — Sharna Wiblen and Alexis Fink, Is HR Tech Undermining Your Strategy? — vendor defaults, satisficing, human glue, static logic, the four-step Strategy-First Check-In — https://sloanreview.mit.edu/article/is-hr-tech-undermining-your-strategy/
- SHRM, 2025 — 2025 SHRM State of the Workplace Research Report — fewer than half of HR professionals and employees regard their company’s HR technology as effective — https://www.shrm.org/content/dam/en/shrm/topics-tools/research/2025-shrm-state-of-the-workplace-research-report.pdf
Discover more from Jeffrey Summers
Subscribe to get the latest posts sent to your email.
