Your firm has been looking at deals for years.
Every one of those deals left a trace: a memo, a model, a set of underwriting assumptions, a recommendation, a decision, and a reason for that decision.
For the deals you closed, real operating outcomes followed. The asset either confirmed your thesis or taught you where the thesis was wrong.
That is not a file system.
That is institutional knowledge.
Most firms treat it like a file system.
That is the waste.
What is actually sitting in the folders
If your firm has been acquiring or managing assets for five years or more, somewhere in your shared drive there are:
- first-pass screening memos;
- underwriting models with your firm's assumptions;
- diligence notes, call logs, and property condition summaries;
- investment committee materials;
- approval records and decision memos;
- passing memos explaining why the firm did not proceed;
- post-close business plans;
- monthly and quarterly property-level reports;
- asset management notes, capex decisions, leasing updates, and variance explanations.
Most of that knowledge is effectively inaccessible on demand.
Ask a simple question:
What were our occupancy recovery assumptions the last time we underwrote a similar hotel in this submarket, and how did actual performance compare?
For many firms, the answer depends on whether someone remembers the deal, remembers the folder, and has time to go looking.
That is not a knowledge system. That is archaeology.
The fragility of tribal memory
Most investment firms carry their institutional knowledge in people.
A senior acquisitions director knows which markets the GP distrusts. An asset manager remembers which operator overpromised on labor savings. A principal remembers the deal that looked perfect in committee but underperformed because the renovation timeline was wrong. A controller remembers which property manager's reports always need extra scrutiny.
That knowledge is valuable. It is also fragile.
When people leave, context leaves with them. What remains is a folder structure, inconsistent naming conventions, and several versions of the same Excel file labeled “Final.”
New team members start closer to zero than they should. They ask questions the firm has already answered. They make assumptions the firm has already tested. They re-learn lessons the firm paid for with real capital.
The firm does not really learn from its history. It archives its history and starts over.
The archive should be connected to live work
This is where most knowledge-base projects go wrong.
A firm decides it wants to “organize the archive.” Someone proposes a database. Someone else proposes a better folder structure. The team agrees the idea is useful. Then the system dies because it requires busy people to maintain a second process after the real process is done.
That will not work.
The archive becomes valuable when it is connected to live workflows.
When a deal-screening skill reads a new OM and T-12, it should also record the extracted facts, assumptions, risks, and decision. When an underwriting support skill populates a model, it should preserve the assumptions behind the output. When an asset management reporting skill processes monthly reports, it should capture actual performance against the original plan.
In other words, recurring workflow skills become the intake mechanism for institutional memory.
The knowledge base is not another data-entry obligation. It is the byproduct of doing the work in a structured way.
That is the layer Suzerand cares about: turn messy deal work into usable deliverables today, while preserving the structured memory that makes the next deal faster and smarter.
What it would mean to actually use that knowledge
Imagine a different workflow.
A new deal comes in. Before the team finishes the OM, the firm can ask:
- How many comparable deals have we seen in this market?
- Which ones advanced to committee, and why?
- Which ones died early, and what killed them?
- What entry and exit assumptions did we use last time?
- Where did actual performance diverge from underwriting?
- Which operators, brands, lenders, or counterparties have shown up before?
- What did we believe about this submarket three years ago, and what changed?
These are not exotic questions. They are the questions experienced investment professionals are already trying to answer mentally when they evaluate a new opportunity.
The difference is that most of the evidence is buried.
The firm has the data. It cannot query the data.
Why this has been hard to solve
This problem is not new.
Firms have tried to solve it with CRMs, shared-drive rules, pipeline trackers, deal databases, and analyst-maintained summaries. Some of those systems help. Most decay over time.
The reason is straightforward: they require humans to do classification and maintenance work on top of their real jobs.
Analysts are already trying to get deals screened, models built, memos drafted, and diligence finished. Asking them to maintain a perfect institutional database after the fact is asking the busiest people in the process to do the lowest-status work in the process.
So the system gets partially maintained, then inconsistently maintained, then ignored.
A better approach is to structure knowledge from the documents and workflows themselves.
That means classifying memos, extracting assumptions from models, linking decisions back to source files, tagging the reason a deal passed or died, and comparing actual performance to underwriting after the asset is owned.
Not magically. Not perfectly. But well enough that a serious implementation can turn dormant archives into usable knowledge infrastructure.
What the technology makes possible
A modern AI workflow can ingest historical deal materials and extract structure from documents that were never built for machine readability.
It can read deal memos and tag the thesis. It can parse underwriting models and pull core assumptions: cap rates, revenue growth, occupancy recovery, exit timing, renovation budgets. It can process passing memos and categorize the reason a deal died: price, basis, physical condition, operator risk, market concern, financing, brand issue.
It can ingest actual asset performance and compare it against original underwriting.
That last part matters most.
A deal archive becomes much more valuable when it does not just answer “what did we think?” but also “were we right?”
Where did the firm consistently underwrite too aggressively? Which expenses were underestimated? Which market assumptions held? Which operator promises proved real? Which deal-killers turned out to be false alarms?
That is the beginning of a living knowledge base.
A proof-shaped example
Suppose a hotel deal comes in with a recovery story: low current occupancy, renovation upside, brand reset, stronger management, and a five-year exit.
The team wants to know whether it has seen this movie before.
Before: an associate asks around, searches old folders, finds two old memos, opens three models, and still misses the passed deal from eighteen months ago that had the most relevant submarket assumptions.
After: the system surfaces prior comparable deals, the original recovery assumptions, actual performance where available, the reason each deal advanced or died, and the exact source files behind the answer.
That does not make the investment decision.
It gives the team a better starting point before it uses judgment.
Why this matters strategically
The advantage is not only speed.
A firm with queryable institutional memory can onboard people faster. It can calibrate underwriting with evidence instead of anecdotes. It can preserve the GP's judgment in a system rather than relying on apprenticeship alone. It can revisit old markets with a better starting point. It can identify where its own assumptions have been systematically wrong.
That is a real operating advantage.
It also changes the value of the firm's track record. The track record is not just a marketing exhibit for LPs. It becomes an internal analytical asset: a source of feedback that improves future decisions.
The firms that build this layer will make better use of their history. The firms that do not will keep paying for lessons and then storing those lessons in folders nobody searches.
The practical gap
This is not automatic.
Archives are messy. Folder structures drift. File names are inconsistent. Final versions are not always final. Some files contradict each other. Some decisions were made in email threads or calls that never became formal memos.
A useful knowledge base needs an intentional structure:
- what fields matter;
- how deals should be categorized;
- how to treat passed deals versus closed deals;
- how to connect original underwriting to actual performance;
- how to preserve source traceability;
- how to avoid turning every detail into noise.
That is implementation work.
The right starting point is not to ingest everything and hope intelligence appears. The right starting point is to pick one high-value knowledge question and structure the archive around it.
For example:
- comparable deal memory for a specific asset class;
- underwriting assumption calibration by market;
- passed-deal reason codes;
- operator and property manager history;
- original underwriting versus actual performance.
Start narrow. Build the structure. Let it compound.
The question for a GP or CIO
If a deal came in today — in your asset class, in a market you know — how long would it take your team to pull everything your firm has ever seen in that market, surface the major assumptions from the last comparable deal, and identify what your underwriting has historically gotten wrong in similar situations?
If the answer is “we would have to ask around,” that is the gap.
Not because the firm did anything wrong. Because until recently, turning historical deal work into usable knowledge required more manual maintenance than most teams could justify.
That is changing.
The knowledge already exists. The question is whether the firm can use it when the next decision has to be made.
Every quarter without that capability is another quarter of deals processed without the full benefit of what came before.
Suzerand helps firms structure institutional knowledge through live workflow skills: deal screening, underwriting support, asset reporting, passed-deal memory, and original-underwriting-versus-actuals. If your historical deal work is still trapped in folders, request a workflow review at suzerand.com and choose one archive question worth structuring first.