“Dark social” is a convenient name for sharing and influence that ordinary web analytics cannot observe directly: a guide forwarded in email, a screenshot sent by text, a link pasted into a private group, or an attorney's explanation discussed on a call.
It is not a hidden attribution channel that turns direct traffic into content ROI. The firm can preserve evidence without pretending to see the whole journey.
Keep the term inside its boundary
Private sharing may hide the original referrer from analytics. Other things create the same symptom: bookmarks, privacy controls, untagged documents, app behavior, copy-and-paste, redirects, device changes, and tracking failures.
Therefore:
- direct traffic is not automatically dark social;
- a missing source is unknown, not organic influence;
- self-reported recall is evidence with memory and prompting limits; and
- a forwarded article may contribute without causing the inquiry.
Use the term to describe an observability gap, not to fill it with a preferred story.
Build content that survives forwarding
A recipient who receives a screenshot or pasted excerpt may lose the page title, firm, date, source, and conditions. Design substantial work so the core idea remains identifiable:
- clear attorney and firm identity;
- descriptive title and date;
- supportable thesis;
- source links and material limits;
- useful section headings and quotable context;
- stable URL and updated/archived status; and
- an accurate next step.
Do not put essential qualifications only in a footer that disappears from the forwarded passage.
A referrer needs more than a firm announcement. A useful resource helps them recognize a boundary, prepare a client, or explain why a conversation may be relevant. Juris Digital's 2021 focus group on how people find a lawyer offers dated qualitative context from 25 participants who discussed recommendations alongside online research. It is not current population evidence or an estimate of private sharing.
For a specialist appellate firm, a guide might explain questions trial counsel should consider before final judgment. For a succession practice, it might help accountants recognize when ownership planning becomes a legal and operating conversation.
Preserve the professional relationship in the record. If an accountant sends the guide and introduces the prospect, content should not take ownership of the referral.
Ask about influence without leading
Use a neutral question:
“Before contacting the firm, what—if anything—did you see, receive, hear, or discuss that helped you decide to reach out?”
Keep the response in the person's own words, then classify named content, attorney, outlet, referral, search, ad, event, prior relationship, none recalled, or unknown. Record who asked, when, and whether the response was prompted.
Do not ask, “Did someone text you our guide?” A yes answer may reflect the question more than the journey.
Use tracking without overstating it
Campaign parameters, short links, QR codes, call tracking, form fields, server logs, and referral records can preserve observable paths. Their absence does not establish no influence; their presence does not establish causation.
Keep these separate:
- observed incoming path;
- neutral reported influence;
- professional referrer;
- platform or vendor attribution;
- distinct inquiry identity; and
- matter stage.
One person can have several influences and one inquiry.
Work a recomputable influence cohort
Consider a fictional business-law guide distributed through attorney LinkedIn posts, a referral email to accountants, and an earned quotation that links to the guide. The firm receives 16 distinct inquiries during the defined theme period. Twelve contain at least one observed or reported theme influence.
The raw flags are:
- six records with an observed guide link;
- seven records that neutrally name the guide or attorney explanation;
- four records with a professional referrer who shared the guide; and
- two explanation-only records with no remembered title or link.
Those flags total 19, but the union is 12 people. Reconcile it:
- start with six observed-link records;
- add three named-guide records with no observed link;
- add one referral-only record with neither link nor named-guide response; and
- add two explanation-only records.
That produces 6 + 3 + 1 + 2 = 12 distinct influenced inquiries. Within the first nine records, three people have all three flags: observed link, named guide, and referrer. One additional observed-link record also names the guide. The three triple overlaps contribute six duplicate flags, and the additional observed-plus-named record contributes one. Those seven duplicates explain why 19 flags resolve to 12 people.
Of the 16 total inquiries, ten qualify, eight reach attorney review, four sign agreements, three open matters, and one remains signed pending opening. Among the 12 influenced records, the firm should not claim all four agreements or three openings were caused by private sharing. It can report the observed/reported paths and stages, then compare fit, timing, alternate influences, costs, and capacity.
The decision is continue the guide theme and repair influence capture. Two intake interviews did not ask the neutral question, and one referral was stored only in free text. The firm corrects those fields, preserves the pending matter, and declines to assign a dollar value to direct traffic.
Report a chain, not a certainty score
An honest report can state:
- what content and distribution existed;
- which links and referral paths were observed;
- what prospects and referrers reported;
- how identities were deduplicated;
- which inquiry and matter stages matured;
- which costs and capacity applied; and
- what remains unknown.
Avoid “dark social generated 40% of cases,” a proprietary influence score without inspectable rules, or adding touch categories as people.
Attorney content still needs accurate claims and responsible reuse. ABA Model Rule 7.1 supplies a model-rule baseline; the firm's jurisdiction and facts control.
Juris Digital's current Solutions hub provides broad context for content, communications, and measurement work. Bring the exact content, distribution records, observed paths, neutral influence responses, referral fields, identity rules, and mature inquiry cohort. Ask for a written scope covering tracking, intake fields, reconciliation, reporting, access, fees, and attribution limits. The hub does not establish a dark-social product or promise recovered attribution or matters.
Last updated:
