What is a well-documented building worth?
A building becomes easier to manage, assess and hand over when the work done to it can be followed. AI gives us better ways to assemble that knowledge and use it while we still own the property.
A building can have been well managed for years and still be difficult for a buyer to assess. The work has been done, the invoices paid, and someone on the owner’s side knows how it all fits together. Yet when the records are examined, there are questions that can only be answered by finding the right person and hoping they remember.
After many viewings of properties and portfolios, we have come to appreciate the difference between being told what has been done and being able to follow it ourselves. An invoice for completed work is useful. When it can also be connected to the right installation, an account of the scope and a record of the follow-up, we have a much better basis for understanding what we are considering buying.
It is difficult to put a separate price on that confidence. It nevertheless affects what we can assess, the questions we need to ask and how much of the property must still be treated as uncertain.
That is why we think documentation deserves a larger place in the discussion about developing a property’s value. We readily spend time working out which investments the building needs. The knowledge that helps us choose those investments, and later shows what the money achieved, deserves similar care.
AI makes the question more interesting. Reading older records, finding relevant information and bringing it together no longer needs to be an entirely manual task. There is an opportunity to go further with the documentation, including in a small business where the person who knows the building best has plenty of other work to do.
It should be possible to follow the work
A well-organised document folder is a useful beginning. It helps us find what has been saved. The standard we are interested in also needs to answer questions about how the information fits together and what it means for the work still to be done.
An inspection report may describe a defect without establishing whether it was subsequently remedied. An invoice may show what was charged while leaving it unclear whether the work covered the whole installation or a small part. A maintenance plan may contain a replacement year without explaining why that year was chosen. Each document can be accurate and still leave the next person with a considerable task of interpretation.
We want to be able to follow the connection from an observation to a decision, then to what was ordered and what was recorded as completed. Where the work covered only part of something, that boundary needs to be understandable too. If somebody later asks why a particular job appears in the maintenance plan, the reasoning should have somewhere to live where it can be found.
That history becomes useful long before anyone starts preparing a sale. A property manager commissioning another job can see what has already been tried. Someone taking over an assignment can understand why a question remains open. At the next budget review, there is a better basis for distinguishing planned work from work that has been completed but still needs following up.
It also reveals something about how the company works. Current, verifiable records let an outsider see evidence of planning and follow-through, rather than relying entirely on the owner’s account. A tidy register cannot establish that a building is well managed, but the ability to examine the claims makes the account considerably more credible.
A buyer needs to assess what remains
Uncertainty has a practical consequence in an acquisition appraisal. If the extent of completed work cannot be established, the buyer needs to decide whether to investigate further, make a more cautious assumption or bring the issue into the negotiation. A seller may have had good control of the property and still struggle to convey that control through the material handed over.
Very good documentation has a clear job here. It can help settle what remains before the buyer needs to allow for a problem that may already have been resolved. That is our judgement about how better information can help a transaction, rather than a fixed discount that applies whenever a document is missing.
Reviewing the records already belongs within the established process. RICS’ standard for technical due diligence of commercial property provides a framework for investigating a building’s condition and risks ahead of a decision. Well-prepared material gives the surveyor more to work with; the investigation and professional judgement are still needed.
There is relevant price research too. A Hong Kong study by Yung Yau and Daniel Chi Wing Ho reported a positive association between good documentation and residential prices. The 2009 abstract mentions as-built drawings and incident records. We could read the abstract, not examine the full model. It supports taking the question seriously, without establishing a Swedish price premium.
The sensible place to begin is with information that can change the assessment. Evidence that a major installation has been replaced may give a buyer reason to reconsider an allowance for future investment. Records showing that only part was replaced give a different picture. Documentation makes the distinction assessable, with the financial consequence arising through the assumptions that change.
Sometimes it also shows that the property is in worse condition than previously thought. An older report may reveal an issue that still needs investigating, or work that was never completed. The information can make an appraisal less favourable while leaving the owner better equipped to make the next decision. For a long-term owner, that is part of the benefit of knowing what one owns.
Adding a general documentation premium on top would move too quickly. If verified work has already changed the appraisal’s expected future expenditure, a further premium needs its own justification. Our conviction about the value of good records does not depend on every financial consequence appearing as a separate line in a valuation.
What the lender can rely on
The same ability to account for the property matters in a conversation with a lender. Our experience is that the discussion improves when we can show what has been done, what we are planning and how we arrived at the cost estimates. Questions can be answered with material the bank can examine, and the conversation can move forward from there.
For a smaller owner, this can be an important way to demonstrate how the business is run. A maintenance plan with explained assumptions, documented investment and work followed through gives the lender something concrete to assess in its relationship with the owner. We see this as ongoing work on trust, including between the occasions when we want to discuss a new loan.
Some finance products make the connection with documentation particularly clear. SEB, for example, requires applicants for its green property loan to show that the building meets its criteria, using evidence such as an energy performance declaration, an energy calculation or certification. The records substantiate a characteristic that can qualify the building for the product.
That is a more specific financial function than a general hope that a bank will appreciate a well-kept archive. For ordinary borrowing, however, we need to distinguish a better conversation from a better credit decision. We have no basis for promising a lower interest rate or higher leverage because of documentation quality alone. We can work deliberately on trust without making that promise.
AI can make the history usable
The ambition can otherwise run into a practical obstacle: examining the records takes time. An older report needs reading, information needs connecting to the right part of the building, and conflicting accounts need finding before anyone starts treating the summary as authoritative. As more documents arrive, the account needs keeping current.
Runs in our operation: we have had an assistant read older property documents and propose information mapped to our system’s structure. It was asked to provide source references and lists of gaps and conflicts, which we then reviewed ourselves. The result could become both information to enter and a list of matters needing further investigation.
That changes the division of work. With the information assembled and referenced, more of our review can concentrate on what it means for the property. An unknown installation year becomes a question to resolve. An older assessment of a façade needs comparing with its condition today. The assistant helps us reach those questions. We still decide what is reliable enough to enter and what needs checking on site.
A simpler way in is The Property Workspace, which we are preparing as a Build in The Method. Tested by us: the pilot brings a property’s documents into an AI project, with instructions, a source register and checks. Its purpose is to produce answers that can be checked against the documents and make missing information visible. It is a lighter setup than our operations system, without automated case handling, and has been tested on practice material.
For us, the next substantial opportunity is to use AI as the records grow. Our assessment is that an assistant could compare a new inspection report with earlier material and propose changes, identify contradictions and list questions for follow-up. That is a more bounded and useful assignment than asking it to decide whether the whole building is in good condition. Proposals need to lead back to the sources and be approved before changing the account the business works from.
The material also needs to be readable and suitable for use in the chosen service. A scan may need text recognition and checking, while personal information needs handling according to the business’s rules. A persuasive summary is of little help if its references lead to the wrong place or old information is presented as the current state of the building.
The next job should leave a better record
Being able to read the archive is a beginning. The larger benefit comes when the next job leaves better information behind as part of the work itself, so that the same context does not need reconstructing when another question arises.
We already run our property operations on v4, which builds on v3 and is still being developed. The building’s parts sit within a shared structure alongside connections between the property’s controls, work orders and supporting evidence. An order preserves what was requested when it was created. This gives us a practical form of digital building twin: a structured account of what we have recorded about the building, available for use in working on it.
We want to understand which requirement a job was carried out against and find the material left afterwards. Protections for the history can help preserve that connection, but they cannot turn an inadequate photograph into a complete account. The people carrying out and reviewing the work still need to ensure that the records say what the next person will need to know.
Our experience is that the connections are now considerably clearer than in the material we inherited. We spend substantial time developing the systems and using AI to improve that account, while the information itself needs extending through the work of managing the buildings. It is an ongoing part of ownership, already useful when we need to understand what work to undertake next.
We therefore see good reason to improve the documentation early after an acquisition. Our knowledge can improve before much has changed in the building, helping us choose where to spend money next. AI offers help with work that might otherwise remain an intention to have somebody go through the binders.
Over time, we want that knowledge to travel with the property when a manager changes, someone else takes over a job and eventually a new owner takes over the building. The next person should be able to spend their time on the next improvement, without first having to discover things we already knew but never managed to hand over.