The Authentic Social Presence

Presence is not reach.

There is a familiar model for building an online presence.

Create content. Get reach. Convert reach into followers. Turn followers into an audience. Eventually, perhaps, turn the audience into a community.

The model is so familiar that it is easy to mistake it for the structure of the internet itself.

But there is another possible developmental path—one that begins before a formal audience exists.

A thoughtful comment. A conversation in a group. A respectful disagreement. A recurring appearance in a discussion. Someone seeing the same name for the third time. Someone else remembering a response from two months ago. A person encountering someone on one platform and later recognizing them on another.

None of those moments necessarily looks significant in analytics.

A comment can receive three likes and still leave a memory. A conversation can produce no new follower and still alter how a person is represented in someone else's mind. A public exchange can be observed by people who never participate in it.

Something may have changed even when the dashboard barely moves.

Someone knows who you are now.

That distinction is the starting point for what I call The Authentic Social Presence: a proposed framework for understanding how durable person-level recognizability can accumulate through socially meaningful encounters before it becomes visible as formal audience membership.

Exposure and presence are not the same variable

Exposure means that someone encountered a stimulus.

Presence, as I use the term here, means that encounter history has accumulated enough socially meaningful information for a person to become a persistent object of recognition and expectation in another person's social world.

Exposure is not socially inert. The classic mere-exposure literature shows that repetition alone can alter familiarity and evaluation. Robert Zajonc's foundational work and Robert Bornstein's later meta-analysis establish that repeated exposure can matter even without rich interaction.

But there is a difference between stimulus familiarity and accumulated person knowledge.

I can recognize a logo because I have seen it many times. I can also recognize a person because I have seen how they speak, what they care about, how they respond under disagreement, what they repeatedly contribute, and how they behave toward other people.

The second form of recognition carries structure.

Exposure:
"I have seen this before."

Accumulated presence:
"I know who this is, and I have expectations about them."

That difference is where this framework begins.

The closest theoretical ancestors

The phrase social presence already has an important academic history. Traditional Social Presence Theory concerns the extent to which another actor is experienced as socially salient, real, or co-present through a medium. The framework proposed here is not intended to replace that tradition.

The distinction is temporal and person-specific. I am interested less in how socially present a medium makes someone feel in a given interaction and more in how a particular person becomes recognizable through accumulated interaction history over time.

Joseph Walther's Social Information Processing perspective is probably the closest theoretical ancestor. Walther argued that people can accumulate individuating information through computer-mediated messages and develop increasingly interpersonal relationships despite reduced nonverbal cues. His longitudinal work with Judee Burgoon reinforced the importance of understanding mediated relationships developmentally rather than as single encounters.

Christian Licoppe's concept of connected presence is another important neighbor: mediated relationships can be maintained through ongoing streams of small exchanges rather than only occasional substantive communication.

Ambient-awareness research comes even closer to one of the core mechanisms proposed here. Work by Anna Levordashka and Sonja Utz found that streams of individually small social-media updates can accumulate into knowledge about other people and feelings of closeness.

None of these literatures, by itself, is identical to the framework I am proposing. But they establish that the underlying mechanism is not invented from nothing.

The interesting question is what happens when we combine accumulated mediated impressions with public observation, reputation, parasocial familiarity, network topology, cross-context identity, and algorithmic recurrence.

The original linear model was too simple

My first formulation looked like this:

Authentic Presence
        |
        v
Recognition
        |
        v
Relationship
        |
        v
Trust
        |
        v
Influence

It is intuitive, but research makes clear that it should not be treated as a universal sequence.

Parasocial relationships can develop without reciprocal interpersonal relationship. Reputation can produce expectations without direct interaction. Social proof and network reinforcement can influence behavior without deep trust. Negative recognition can become extremely strong. Institutional authority can create influence before familiarity. Virality can produce immediate behavioral effects. Influence can also create new exposure, which feeds back into the process.

The stronger model is probabilistic, recursive, and network-conditioned:

Encounter history
        |
        v
Recognition
        |
        v
+---------------------------+
| Relational knowledge      |
| Parasocial familiarity    |
| Reputation                |
+---------------------------+
        |
        v
Trust / distrust / credibility / expectancy
        |
        v
Context-dependent influence
        |
        v
Behavior / sharing / response
        |
        +-------------> new encounters

This matters because it changes the theory from a success pipeline into a state system with bypasses, reversals, decay, and feedback.

Person-specific relational memory

One of the central mechanisms is memory.

Earlier I used the phrase relational memory, but that term already carries technical meanings in cognitive science. A more precise phrase here is person-specific relational memory or accumulated person knowledge.

The idea is simple: previous encounters leave residual information that changes later encounters.

That information can include identity recognition, remembered behavior, prior exchanges, topical associations, attributed dispositions, expectations, and knowledge of how the person behaved around others.

A conceptual memory-stock model might look like this:

M(i,j,t) = retention * M(i,j,t-1)
           + sum(weighted encounters)

The equation is not offered as an established psychological law. It makes explicit a more modest claim: encounters differ in informational weight, memories decay, and new encounters update what is already known.

That accumulated state is why the tenth encounter with a recognizable person is not cognitively equivalent to the first.

Weighted encounter density

Raw frequency is not enough.

Ten passive impressions should not automatically be treated as equivalent to ten substantive exchanges.

A more useful construct is weighted encounter density: the frequency of encounters over time adjusted for their social informativeness.

Possible weights include:

• interaction depth,
• context diversity,
• reciprocity,
• public observability,
• emotional salience,
• identity consistency,
• source credibility,
• and whether the encounter confirms or violates prior expectations.

This also connects the framework to network science. Research on complex contagion shows that diffusion depends not only on exposure count but on where reinforcement comes from and how the network is structured. One encounter through a strategically positioned or trusted intermediary can matter more than many undifferentiated impressions.

Presence is therefore not just a quantity problem.

It is a history-and-topology problem.

Observed participation is first-class social evidence

One of the most important extensions of the framework is that direct interaction is not required for person knowledge to accumulate.

Imagine repeatedly seeing the same person participate in a Facebook group.

You watch them disagree with someone. Then you see them encourage someone else. Later they ask a careful question. You see them admit an error. You notice how they respond when challenged. You have never spoken to them.

But you have learned something about them.

This is not merely content consumption. It is observation of behavior in social context.

Creator model:
Person -> Content -> Viewer

Observed-participation model:
Person -> Other person -> Response -> Observer

Research on reputation and indirect reciprocity provides a broader mechanism for this. Information about how someone behaves can affect how third parties evaluate and treat them even without direct experience.

That means a public comment is not only communication with its recipient.

It can also become reputational evidence for everyone watching.

In this framework, observed participation is first-class evidence.

Known before followed

Platforms make formal affiliation visible.

Follow. Subscribe. Connect. Friend. Join.

Because those transitions are measurable, they are often treated as the beginning of the relationship.

But the human sequence may have started much earlier.

Encounter
   |
Encounter
   |
Recognition
   |
Observed participation
   |
Conversation
   |
Return encounter
   |
FOLLOW

The platform records the final transition.

The person carries the entire history.

I call this the Known Before Followed hypothesis:

Measured recognition and accumulated encounter history may sometimes predict future formal following better than raw impression count alone.

That is not something I want to assume. It is something that can be tested.

The latent recognition network

This hypothesis leads to a larger idea.

A follower graph is not necessarily the same thing as a recognition network.

Someone can recognize you, remember you, quote you, distrust you, recommend you, criticize you, transmit an idea associated with you, or know how you tend to behave without ever clicking Follow.

I call the population of such people a latent recognition network.

Latent Recognition Network != Follower Graph

This distinction may be one of the framework's most useful consequences.

Imagine two creators.

One has ten thousand followers but is weakly remembered by most of them.

Another has two thousand followers but thousands of additional people scattered through communities, comment sections, forums, professional spaces, and multiple platforms who recognize the person when they appear.

Which person has the larger social presence?

Follower count cannot answer that question by itself.

The latent recognition network is still a research proposition. But if it can be measured, it may explain forms of diffusion and reputation that platform-native audience metrics miss.

Cross-context identity integration

The internet makes the problem even more interesting because human memory crosses boundaries that analytics do not.

Someone may first encounter a person on Facebook, later see a YouTube video, then notice the same name on LinkedIn, and months later read an essay on a personal website.

Those platforms do not necessarily share a complete representation of that social history.

The observer does.

But continuity should not be assumed. Research on self-presentation shows that people often present themselves differently across platforms and audiences. Context collapse creates different pressures in different social environments.

So the stronger construct is not cross-platform continuity but cross-context identity integration:

Cross-context encounters
          |
          v
Can the observer bind them to the same person?
          |
     +----+----+
     |         |
    yes        no
     |         |
     v         v
Integrated   Segmented or conflicting
representation representation

This is one of the most promising open empirical questions in the framework.

Does encountering the same recognizable person across several contexts produce stronger or more durable person-level recognition than an equivalent number of encounters in one context?

The answer should not be assumed.

It should be measured.

Algorithms are selection–transformation–feedback infrastructure

My first instinct was to describe the algorithm as a transport layer.

That is useful, but incomplete.

Algorithms certainly affect distribution. They help determine who encounters whom, how often, in what sequence, and under what ranking conditions.

But contemporary algorithmic systems can also transform the encounter itself.

Experimental work on AI-generated smart replies has shown that AI assistance can alter conversational language and interpersonal evaluations. Research on algorithm responsiveness also examines how people interpret the degree to which recommendation systems seem to recognize or respond to their identities.

A better description is therefore:

Algorithms are selection–transformation–feedback infrastructure for social presence.

Selection
  -> changes encounter probability and sequencing

Transformation
  -> changes what is said, surfaced, suggested, or perceived

Feedback
  -> user behavior changes future selection

The system becomes recursive:

Actor behavior
      |
      v
Platform selection
      |
      v
Encounter
      |
      v
Attention + person memory
      |
      v
Recognition / expectation
      |
      v
Response / follow / share / ignore / challenge
      |
      +---------------------> platform selection again

The narrower part of the original intuition survives: an algorithm cannot guarantee earned interpersonal trust simply by producing repeated exposure.

But it can shape the histories and signals from which human judgments are formed.

Authenticity is a moderator, not magic

The word authentic in the framework title needs care.

We cannot directly inspect another person's inner motives. Online authenticity is perceived, inferred, performed, negotiated, and sometimes strategically managed.

So authenticity should not be treated as an objective substance that automatically produces influence.

A better formulation is perceived authenticity: the observer's inference that a person's behavior is sufficiently consistent, sincere, or non-instrumental to be treated as genuine.

This matters because repeated encounters create repeated opportunities to update that inference.

If every conversation appears to be a disguised conversion opportunity, accumulated interaction can weaken trust rather than strengthen it.

If someone behaves coherently across many settings—not identically, but recognizably—the observer receives evidence that the represented person persists across contexts.

The theory therefore does not say:

authenticity -> influence

It says perceived authenticity can change how encounters are interpreted and how strongly they update person knowledge, trust, distrust, and expectancy.

Recognition is valence-neutral

This is crucial.

Becoming recognizable is not the same thing as becoming liked.

A propagandist, respected expert, community organizer, troll, helpful neighbor, critic, entertainer, and adversary can all become highly recognizable.

Recognition strength and evaluative valence must therefore remain separate variables.

R = recognition strength
V = evaluative valence

High recognition can coexist with negative valence.

Trust can become distrust. Relationship can become distance. Familiarity can become annoyance. A reputation can become a warning signal.

That does not break the model.

It reveals why the model must describe social legibility before it describes positive influence.

A state-transition view

One way to make the framework operational is to treat social recognition as a partially ordered, reversible state system rather than a ladder.

Unknown
  |
  v
Exposed
  |
  v
Recognized
  |\
  | \-----------------> Reputation-known
  |-------------------> Parasocially situated
  |
  +-------------------> Relationally situated
                           |
                           v
                    Trust / distrust
                           |
                           v
                       Influence
                           |
                           v
                 Share / respond / follow
                           |
                           +----> new encounters

Possible reversals:
Recognized -> Forgotten
Trusted -> Distrusted
Relationally situated -> Distanced

Followed is deliberately not treated as a developmental state.

It is a platform action that can occur almost anywhere in the sequence.

Someone can follow after one viral exposure. Someone else can follow after months of recognition. Another person can know and trust a creator for years without formally following at all.

That is exactly why audience membership, recognition, trust, and influence should be measured independently.

What would make the theory scientifically interesting?

A framework becomes more useful when it risks being wrong.

The following propositions could be tested:

1. Meaningful Encounter Advantage: holding total impressions constant, socially informative encounters should produce greater delayed person recognition than passive exposure.
2. Encounter Diversity Effect: distinct contexts should improve recognition relative to repetitive identical exposure when identity cues remain bindable.
3. Cross-Context Integration Effect: multi-platform or multi-community encounters should improve person recognition only when observers can successfully bind the encounters to the same identity.
4. Observed-Behavior Reputation Effect: repeatedly observing someone's interactions with third parties should alter trustworthiness judgments even without direct interaction.
5. Known-Before-Followed Effect: measured recognition should predict later following better than impression count after controlling for prior affinity and content evaluation.
6. Latent Recognition Network Effect: recognizable non-followers should predict later diffusion beyond follower count alone.
7. Authenticity Moderation Effect: encounter history should relate more positively to trust when interactions are perceived as non-instrumental.
8. Recognition–Valence Separation: encounter frequency should be capable of increasing recognition while liking or trust decreases.
9. Network Reinforcement Effect: equal exposure counts should produce different effects depending on network topology and source overlap.
10. Recognition Decay Effect: person-specific memory should decline without reinforcement, with high-salience and diverse encounters decaying differently from passive impressions.
11. Violation Asymmetry Effect: a salient violation after substantial trust accumulation may produce a larger negative update than an ordinary positive encounter produces a positive one.
12. Influence Feedback Effect: once a person generates measurable influence, resulting comments, shares, and recommendations should increase future encounter probability.

These propositions separate the framework from the vague claim that relationships matter.

They turn it into a research program.

What would weaken or falsify it?

Several findings would force revision.

If meaningful repeated encounters produce no stronger person recognition than equivalent passive exposure, the framework's most important distinction would weaken.

If observed participation contributes almost nothing to impression formation, that mechanism should be removed or narrowed.

If recognition does not meaningfully predict later following after prior affinity and content quality are controlled, the Known Before Followed hypothesis would lose importance.

If cross-context encounters rarely integrate into unified person representations, cross-context identity integration would be a niche effect rather than a central one.

If latent recognition networks do not predict diffusion beyond formal follower graphs, they may be cognitively real but operationally unimportant.

And observational research must always confront a major confound: perhaps people interact repeatedly because they already share interests or affinity. Apparent influence can be selection rather than a causal effect of presence.

Any serious empirical work would need to separate:

Presence -> Influence

from

Prior affinity -> more encounters
Prior affinity -> greater susceptibility

That is why controlled experiments, longitudinal designs, network studies, and careful causal inference matter.

The audience may be the visible surface of something larger

This returns us to the original practical observation.

We normally imagine an audience as the network.

Perhaps an audience is only the formally visible portion of a larger recognition structure.

Underneath it may sit weak ties, familiar strangers, past interlocutors, silent observers, community members, people who recognize a face, people who recognize a writing style, people who remember an argument, people who saw someone else respond to you, and people who have encountered you enough times to carry an expectation about what kind of person you are.

Some will eventually follow.

Some never will.

Some may still transmit your ideas.

Some may mention you to someone else.

Some may challenge you publicly and thereby increase recognition among observers.

Seen this way, building an online presence is not simply the accumulation of an audience.

It is the gradual formation of a social topology.

From content graph to person graph

Most online analytics are organized around content objects.

A post. A video. An article. A comment. A thread.

Each object receives its own numbers.

But from the observer's perspective, these objects can gradually collapse into a representation of a person.

Post A
+ Comment B
+ Video C
+ Conversation D
+ Article E
-----------------
"I know who this person is."

The unit of interpretation changes.

At first, the observer evaluates content.

Later, the content arrives from a recognized source carrying accumulated history.

The next message no longer begins from zero.

That is the central mechanism of the framework.

Presence as a distributed state

Maybe an online presence is not best understood as something a person possesses.

Maybe presence is a state distributed across other people's minds and across the network structures that make future encounters possible.

You do not own recognition.

Other people perform it.

You do not fully own reputation.

Other people carry it.

You do not possess social presence in the same way you possess a domain name, account, or mailing list.

Presence exists because representations of you have become distributed throughout a social environment.

Part of that representation may consist of tiny fragments:

"I've seen him before."

"I remember what she said last time."

"I disagree with him, but I know where he's coming from."

"I've watched how she treats people in that group."

"That's the same person I saw on YouTube."

Each representation is incomplete. Some are inaccurate. Some are negative. Some decay. Some are updated by later encounters.

But collectively they constitute something that conventional audience metrics do not directly measure.

Presence.

Presence before influence

The final claim should be stated carefully.

Influence does not always require prior relationship or trust.

Presence does not guarantee influence.

Recognition does not guarantee affection.

And algorithms do more than transport human interaction.

But one important pathway to durable networked influence appears to be this:

Repeated socially informative encounters
               |
               v
Accumulated person-specific memory
               |
               v
Recognition
               |
               v
Relational / parasocial / reputational representation
               |
               v
Trust, distrust, credibility, expectancy
               |
               v
Context- and network-dependent influence
               |
               v
Behavior, transmission, and new encounters

The novelty is not the claim that repeated interaction creates familiarity. That is already well established.

The proposed contribution is the synthesis: person-level recognizability as a history-dependent state that can accumulate through direct interaction, observed participation, ambient exposure, reputation, multiple communities, and potentially multiple platforms before becoming visible as a conventional audience.

That is why I continue to find the phrase Presence Before Influence useful.

Not as a law.

As a characteristic pathway.

Algorithms can accelerate, suppress, select, and transform encounters.

They do not eliminate the accumulated social history through which people become known.

Reach can disappear in seconds.

A post moves through a feed and is gone.

But sometimes an encounter leaves something behind.

A name.

A memory.

An expectation.

A relationship.

A reputational trace.

A node in somebody else's social world.

And perhaps that is what an authentic social presence actually is.

Research foundations and adjacent literature

The framework above was developed in conversation with established research rather than as a claim that its component mechanisms are unprecedented. The following sources are among the most important theoretical neighbors.

1. Short, J., Williams, E., & Christie, B. (1976). The Social Psychology of Telecommunications. Wiley. Foundational Social Presence Theory lineage.
2. Walther, J. B. (1992). "Interpersonal Effects in Computer-Mediated Interaction: A Relational Perspective." Communication Research. doi:10.1177/009365092019001003
3. Walther, J. B., & Burgoon, J. K. (1992). "Relational Communication in Computer-Mediated Interaction." Human Communication Research. doi:10.1111/j.1468-2958.1992.tb00295.x
4. Licoppe, C. (2004). "Connected Presence: The Emergence of a New Repertoire for Managing Social Relationships in a Changing Communication Technoscape." Environment and Planning D: Society and Space. doi:10.1068/d323t
5. Levordashka, A., & Utz, S. (2016). "Ambient Awareness: From Random Noise to Digital Closeness in Online Social Networks." Computers in Human Behavior. doi:10.1016/j.chb.2016.02.037
6. Zajonc, R. B. (1968). "Attitudinal Effects of Mere Exposure." Journal of Personality and Social Psychology. doi:10.1037/h0025848
7. Bornstein, R. F. (1989). "Exposure and Affect: Overview and Meta-Analysis of Research, 1968–1987." Psychological Bulletin. doi:10.1037/0033-2909.106.2.265
8. Horton, D., & Wohl, R. R. (1956). "Mass Communication and Para-Social Interaction." Psychiatry. doi:10.1080/00332747.1956.11023049
9. Granovetter, M. S. (1973). "The Strength of Weak Ties." American Journal of Sociology. doi:10.1086/225469
10. Centola, D. (2010). "The Spread of Behavior in an Online Social Network Experiment." Science. doi:10.1126/science.1185231
11. Serva, M. A., Fuller, M. A., & Mayer, R. C. (2005). "The Reciprocal Nature of Trust." Journal of Organizational Behavior. doi:10.1002/job.331
12. Marwick, A. E., & boyd, d. (2011). "I Tweet Honestly, I Tweet Passionately: Twitter Users, Context Collapse, and the Imagined Audience." New Media & Society. doi:10.1177/1461444810365313
13. Audrezet, A., de Kerviler, G., & Guidry Moulard, J. (2020). "Authenticity Under Threat: When Social Media Influencers Need to Go Beyond Self-Presentation." Journal of Business Research. doi:10.1016/j.jbusres.2018.07.008
14. Davidson, B. I., & Joinson, A. N. (2021). "Shape Shifting Across Social Media." Social Media + Society. doi:10.1177/2056305121990632
15. Hohenstein, J. et al. (2023). "Artificial Intelligence in Communication Impacts Language and Social Relationships." Scientific Reports. doi:10.1038/s41598-023-30938-9
16. Taylor, S. H., & Choi, M. (2022). "An Initial Conceptualization of Algorithm Responsiveness." Social Media + Society. doi:10.1177/20563051221144322