Hello — I’m the Harborfield Assistant. I can explain Harborfield, compare reports, or help you find a relevant starting point. Final product access is determined separately through eligibility.
Loading
Preparing the requested page…
Loading
Preparing the requested page…
Hello — I’m the Harborfield Assistant. I can explain Harborfield, compare reports, or help you find a relevant starting point. Final product access is determined separately through eligibility.
InsightsResearch Method
Many weak business analyses do not fail because every fact is wrong.
They fail because facts, assumptions, and decisions are blended together.
A market observation becomes a prediction. A personal expectation becomes a conclusion. A chosen action is presented as though the evidence forced it.
Separating these categories produces clearer reasoning.
Consider:
Five comparable providers publicly offer fixed-scope podcast-editing packages.
Customers probably prefer fixed packages because they are easier to understand.
I will test a fixed package before considering hourly pricing.
These statements serve different purposes.
It describes an observable public fact within a defined research sample.
It proposes an explanation or expectation that has not been established by the evidence.
It records an action selected using evidence, assumptions, constraints, and judgment.
A strong analysis preserves those labels.
An evidence register does not need to be complicated.
For each important observation, record:
For example:
Evidence statement: Four of seven comparable providers publish a starting package price.
Limitation: The sample is non-exhaustive, and the providers' scopes are not identical.
That limitation belongs inside the evidence record, not in an afterthought at the end.
Assumptions should be written as assumptions.
Examples:
Assumptions can then be classified.
If wrong, they could materially change the decision.
If wrong, they may change details without changing the primary conclusion.
This distinction helps determine what deserves additional research.
A decision should not emerge mysteriously from a report.
State the criteria.
For example:
Choose the positioning option with the strongest combination of prior-experience fit, clear service scope, usable comparable-provider evidence, and manageable operating complexity.
Now the reader can understand why particular evidence matters.
Without criteria, additional information does not necessarily improve the decision.
Assumptions often appear through ordinary language.
“The market is growing.”
Growing according to which measure?
“Customers want simpler packages.”
What evidence establishes that preference?
“Competitors charge about $1,000.”
Which providers? For what scope? Based on how many observations?
“This market has less competition.”
According to what observable measure?
Any of these statements might eventually be supportable.
First they need to be decomposed.
Four labels are enough for many research notes.
E — Evidence
Source-supported observation.
A — Assumption
Unverified input or interpretation.
U — Uncertainty
Important issue that cannot currently be resolved.
D — Decision
Action or conclusion selected by the decision-maker.
For example:
E: Six strong comparable providers were identified.
E: Four specialize by customer type.
A: Customer-type specialization may improve message clarity.
U: Public information does not reveal whether specialization improves sales performance.
D: Develop one specialized positioning option and compare it with two broader alternatives.
The final action no longer masquerades as a fact.
Numbers often create a false impression of certainty.
Suppose a solo professional assumes:
The arithmetic is straightforward:
80 / 10 = 8 units of theoretical delivery capacity.
That calculation can be exactly correct while the model remains uncertain.
The available-hours assumption may be wrong. Delivery time may vary. Eight units of theoretical capacity does not mean eight projects will be purchased or completed.
A separate pricing assumption can be compared against that capacity without converting the exercise into a projected monthly revenue or earnings total.
Exact arithmetic does not make uncertain inputs certain.
Some research questions produce legitimate conflicts.
Instead of forcing a single story, preserve the structure.
For example:
Evidence supporting Option A
Evidence supporting Option B
The contradiction is not a research failure.
It may reveal the real trade-off.
After the analysis, produce a short decision record.
Decision:
Proceed with Positioning Option B for the next test period.
Primary evidence:
Strong fit with prior work and a clearly defined deliverable.
Main assumptions:
Customer specialization will not materially reduce relevant opportunities.
Known limitations:
Public pricing evidence is weaker than for Option A.
Review trigger:
Reassess after additional comparable-provider evidence or direct operating experience becomes available.
That record is much more useful than:
Option B is best.
Research supports decisions.
It does not eliminate judgment.
Two professionals can review the same evidence and make different rational choices because:
The purpose of evidence is to make the decision more informed and more transparent—not automatic.
Keep three questions separate:
What do we know?
What are we assuming?
What are we deciding?
Then add a fourth:
What remains uncertain?
That small discipline improves competitor analysis, market research, pricing scenarios, geographic comparisons, and operating plans.
Harborfield reports are structured to distinguish source-linked observations from assumptions, decision criteria, and limitations. The Market Evidence Report applies that discipline to focused market questions, while the Complete Solo Practice Research Dossier uses it across a broader research package. Missing evidence is not converted into certainty.