Reading and Validating a ProspectAI Report

A ProspectAI report is a full research brief on one person, assembled from public and third-party sources. It has eighteen sections, and most of the time you will only need a few of them. Every value links back to the source it came from, which means you can check any claim and curate the report: accept what is right, correct what is wrong, work through the sources behind a datapoint, and restore anything to the way the research first produced it.

We recommend that you treat a report as a researched starting point to verify rather than a finished answer. Prospect research is repeated out loud, since a capacity estimate ends up in a briefing note, an employer ends up in a solicitation, and a board seat ends up in a conversation with the donor. Knowing where a value came from therefore matters as much as knowing what it says.

Note: Edits are organization-wide. If you change a value, your whole team sees the changed version.

The eighteen sections

Section	What it gives you
Executive summary	The whole picture in a few sentences. Read this first.
Profile Summary	The prospect scored across capacity, affinity, recency, and fit.
Capacity estimate	A five-year giving capacity, with the inputs behind it: estimated net worth, real estate, shares, salary range.
Affinities	The causes and issues this person visibly engages with.
Roles and education	Employment, directorships, board seats, and qualifications.
Charitable giving history	Recorded gifts to other organizations.
Giving Summary visualization	That history as a chart, so scale and pattern are visible at a glance.
Contact details	Addresses, emails, and phone numbers found in public sources.
Conversation starters	Openers written for this specific person, drawn from the findings above.
Connections	Relationships found between this person and Your Network or their own contacts.
Similar Profiles	Other people with a comparable profile, usually of similar names. Summaries are free to read.
Volunteering and Advocacy	Unpaid contribution, which is often a better indicator of affinity than giving is.
Recommended Outreach	A suggested approach based on everything the report found.
Due diligence check	A public-source scan for controversy and reputational or legal risk.
Key Readings	The most useful sources, plus any flagged as needing clarification.
Data Summary	What was found, condensed.
Notes	Notes recorded against this report.
Full source list	Every source used, at the bottom, and editable.

Read the executive summary first, then jump to whichever sections your next conversation depends on.

Gift capacity calculation

The capacity estimate is the figure most often quoted in meetings. It is a five-year figure, not an annual one, and it is the higher of two calculations:

gift capacity (over 5 years) = the higher of 5% of net worth, or 10% of annual income

Dataro takes the higher of the two because the inputs behave differently for different donors. A retired donor may have substantial assets and little income, while a senior executive may earn a great deal and hold few visible assets, so using only one measure would understate one of them.

Where visible assets cannot be found, net worth is estimated from income instead:

estimated net worth = 3 × annual income   (when visible assets are unknown)

This multiple is a conservative proxy rather than a figure found in a source, so we recommend that you treat a capacity estimate built on it as an order of magnitude rather than a precise number.

A blank capacity field

If the capacity field is empty, public sources did not yield enough information to estimate from. It does not mean the person has no capacity.

Wealth is unevenly visible: company directors, property owners, and public figures leave public records, while people whose assets sit in private structures often do not. Screening out every blank would remove some of your best prospects, so we recommend that you read a blank as an unanswered question and judge the person on affinity, roles, and connections instead.

The Due Diligence Check

The due diligence check scans public sources for controversies, negative news, and reputational or legal risk. It is there so that you find out about a problem early.

There is an important limit. It is not a structured sanctions or PEP screen. It does not query sanctions lists, politically exposed persons registers, or court records in the systematic way a compliance tool does. If your gift acceptance policy requires a formal screen at a certain gift level, you still need the formal screen. This check is a first pass, not a substitute for a formal screen.

The check also reflects the moment the report was run, so we recommend that you re-run the report before soliciting on research that is several months old.

Confidence signals

Every datapoint in a report carries a confidence signal: a red, yellow, or green bar next to the source citation. It tells you how well-supported that specific claim is, so you know which values you can use as they are and which ones to check before you rely on them.

Signal	Meaning	What to do
Green	Well-sourced, high confidence	Use it. Spot-check if it is going into a solicitation
Yellow	Some ambiguity in the source or the identity match	Verify before acting on it
Red	Needs review	Do not rely on it until you have checked the source

The signal exists because identity resolution can attach a claim to the wrong person when a source is vague, for example a news article naming "R. Patel, a local business leader" with no employer or city attached. If every value in a report looked equally certain, you would have no way to know where to spend your verification time. The signal is therefore a triage tool: it tells you which datapoints to check first when you only have a few minutes before a meeting.

You see the signal in three places:

Next to each datapoint in the body of the report
In the source popup, when you click a datapoint to see where the value came from
At section level, rolled up from the individual claims in that section

The three placements answer three different questions: whether a value is safe to use, why it is rated the way it is, and whether there is anything in the section worth checking.

Datapoint Actions

A datapoint is a summary the model writes from the sources it found, sometimes layered with additional logic or calculations, for example a conversation-starter bullet or a capacity estimate. Each category in a report holds many datapoints.

Click a datapoint in the report and a popup opens with four options:

Action	What it does	When to use it
Accept	Marks the datapoint as verified, and the report then shows it as verified	You have checked the sources and the datapoint holds up
Override	Lets you edit the datapoint directly and save your own version	You know the correct value, for example from a conversation with the donor
Edit sources	Opens the sources behind the datapoint so you can review each one's confidence, red, yellow, or green, and either select or delete an individual source. Deleting here removes that one source, not the datapoint	A source is weak, or refers to a different person
Restore	Reverts the datapoint to its original state, as it was first created	You want to undo your changes and go back to what the research produced

There is no action that removes a datapoint itself: deleting inside Edit sources removes one source from the datapoint, and the datapoint stays in the report.

Override and Edit sources are easy to confuse. Override changes what the report says, while Edit sources changes what the report cites. For example, a donor who has told you they have retired needs an override, and a correct job title backed by a ten-year-old press release needs its sources edited.

An overridden datapoint carries an "Edited" marker, so a colleague reading the report several weeks later can tell which values came from a person rather than from the research. If you change your mind, Restore puts the datapoint back to the state it was first created in.

If a datapoint is supported by more than one source, Edit sources lets you work through them individually: select the two that clearly refer to your donor, delete the one that refers to a namesake, and leave the datapoint intact.

To work through a section:

Review the section and datapoints
Click it to open the source popup and read the citation.
Accept the datapoint, override it with your own version, edit its sources, or restore it, depending on what you find.

The Full Source List

At the bottom of every report is the complete list of sources reviewed, and it is editable.

To correct the sources behind a report:

Read down the sources and accept the ones that are right.
Delete any that are not this person. This happens with common names, and removing a wrong source removes its influence on the analysis.
Add your own source URL if you know something the search missed, e.g. a local news piece, a foundation annual report, or a conference bio.
Click Regenerate report. The analysis re-runs with your corrections applied.

We recommend that you add your own sources. If you hold a giving history, a news article, a foundation filing, or a conference bio that ProspectAI did not find, add the URL and regenerate. The report is rebuilt with that source included rather than keeping it alongside as a note.

Regenerating does not affect your credit balance. It does clear your manual edits, so note anything you want to keep before you regenerate, and add any source URL you want included before you regenerate rather than afterwards.

Key Readings

Reports also include a Key Readings section, which lists the top sources for the prospect alongside sources that need clarification. It is the fastest way into a long report: read the top sources to understand who this person is, then look at the flagged ones to see where the research is less certain.

Organization-wide Edits

All edits apply at the organization level. If you edit a value, your entire team sees the edited version. The same is true of Favorites.

Correcting a wrong employer fixes it for everyone, so the same mistake is not rediscovered by three colleagues in three separate meetings. The other side of that is that a speculative edit becomes your organization's version of the record.

We recommend that you edit when you know a value is wrong, not when you suspect it. If you are unsure, leave the value alone and use the confidence signal instead.

Frequently Asked Questions

A fact in the report is clearly wrong. What do I do?

Open the datapoint and check the cited sources first. It is almost always a shared-name mix-up, which the sources will show you immediately. Use Edit sources to delete a source that belongs to someone else, or Override to set the correct value yourself. If the problem is systematic, contact your CSM with the donor ID and the field.

Is the due diligence check a sanctions screen?

No. It scans public sources for controversies, negative news, and reputational or legal risk. It is not a structured sanctions, PEP, or court-record screen, and it should not be used to satisfy a compliance requirement that calls for one.

How does Dataro know what my donor gave to other organizations?

From public giving records and news coverage, e.g. honor rolls, annual reports, foundation filings, and media mentions. Anonymous gifts leave no public record, so they are not captured. Absence of recorded giving elsewhere is not evidence that none happened.

Can I change a confidence signal?

Not directly, but you can resolve the underlying issue. Accepting a datapoint after checking its sources, using Edit sources to delete a weak or misattributed source, or overriding the datapoint with the correct value all address what the signal is reporting.

Will regenerating a report wipe my edits?

Yes. Regenerating reprocesses the prospect, and your manual edits are not carried over, so note anything you want to keep before you regenerate.