In 2018 Sergey Gorbatov, Svetlana Khapova and Evgenia Lysova published in Frontiers in Psychology the first interdisciplinary systematic review of the personal branding literature. They examined the academic output across three disciplines and their conclusion is clear: the field is dominated by conceptual and prescriptive work, there’s no shared definition, no validated measurement instrument, and the empirical base is thin.
Translated: almost none of the personal branding advice you read has ever been tested.
That doesn’t mean reputation doesn’t matter. It means that to understand what works you need a theory that has been tested, and one very solid one exists.
The theory that explains which signals work
In 1973 Michael Spence published in the Quarterly Journal of Economics the work that would earn him a Nobel: a model of the labour market as a signalling problem. The employer can’t directly observe a candidate’s productivity, so they rely on observable signals.
The condition for a signal to carry information is a single one, and it’s brutal: it must be more expensive to produce for someone who lacks the quality being signalled. A qualification works as a signal not for its content, but because it’s harder to obtain for someone less able. A signal anyone can produce at the same cost transmits no information: it dissolves into the average.
Apply the criterion to the standard personal branding advice:
| Signal | Cost for someone without the quality | Informational value |
|---|---|---|
| A polished profile headline | Zero | None |
| A professional photo | Low | Near zero |
| Motivational posts about your industry | Low | Near zero |
| A public project that can be inspected | High | High |
| A technical explanation that survives questions | High | High |
| People who vouch for you by name | Very high | Very high |
The first three are what 90% of content on the subject talks about. The last three are what moves the decision.
How much your online presence really matters: the experimental data
There are very few field experiments, and one deserves citing. Alessandro Acquisti and Christina Fong published in 2020 in Management Science a study in which they sent applications to around 4,000 US employers, manipulating the information visible on fictitious candidates’ social profiles.
The main result is more sober than you’d expect: average effects on callback rates were small. But a measurable effect emerged for specific subgroups: a candidate identified as Muslim received fewer callbacks in more politically conservative areas.
Two practical conclusions, both uncomfortable:
- Your social profile matters less than the courses tell you.
- When it does matter, it often works through information you wouldn’t want to be relevant.
The first line of defence isn’t posting more. It’s controlling what’s visible.
Self-promotion works worse than you think
There’s an experiment worth more than a lot of theory. Irene Scopelliti, George Loewenstein and Joachim Vosgerau published in 2015 in Psychological Science a series of studies comparing what self-promoters predicted with the actual reactions of listeners.
The result: self-promoters overestimate the positive emotions they’ll generate (interest, admiration) and underestimate the negative ones (annoyance, irritation). It isn’t that self-promotion never works; it’s that there’s a systematic perception gap between speaker and listener, and it runs against the speaker.
Professional reputation isn’t what you say about yourself. It’s what someone else is willing to say about you when you’re not in the room.
And on that there’s solid data: Stephen Burks, Bo Cowgill, Mitchell Hoffman and Michael Housman published in 2015 in the Quarterly Journal of Economics an analysis of firm data from nine companies, finding that workers hired through employee referrals stay longer and are more profitable for the firm. A referral is an expensive signal, because whoever refers you puts their own reputation at stake, and that’s why it carries weight.
How to build a reputation people can verify
- Clean up before you build. Check what’s public on your profiles. In Acquisti and Fong’s experiment the damage came from unintentionally visible information, not from a weak headline.
- Produce an inspectable artefact. A case study, an analysis, a tool, a public document. It’s the only kind of signal Spence’s model predicts will work, because it can’t be faked at zero cost.
- Be specific to the point of seeming narrow. “Marketing” isn’t a position. “Email marketing for clothing e-commerce” is. A generic signal is indistinguishable from noise.
- Replace the claim with the demonstration. Not “I’m analytical”, but a public analysis anyone can check. Same principle as how to write a CV.
- Invest in people who can vouch for you. In the data, referral is the strongest signal. It’s built by working well with someone, not by posting. I’ve written about it in professional networking.
- Dial down direct self-promotion. The gap Scopelliti measured says the tone that feels right to you is already above the level others tolerate.
- Update at low frequency and high density. One useful piece a quarter is worth more than three posts a week nobody re-reads.
My take: what counts is doing hard things in public
I think the phrase “personal branding” itself pushes in the wrong direction, because it suggests the work consists of packaging. The strongest professional reputations usually don’t come from personal branding: they come from doing hard things in public, with the reputation arriving as a side effect.
I also suspect much of the advice industry on this topic is self-referential: the best way to sell personal branding consulting is to have built a personal brand by talking about personal branding. It’s a closed loop that produces examples that don’t transfer to people doing other jobs.
Going on instinct: visibility without substance isn’t neutral, it’s negative. It exposes you to a bigger audience at exactly the moment you have nothing that survives scrutiny. Better to be findable by a few right people than visible to many wrong ones.
Frequently asked questions
Do I have to post on LinkedIn? No. You have to be findable and verifiable. If your field lives elsewhere (a repository, a professional register, a portfolio, a client network), that’s the channel that counts.
How long does it take to see results? Much longer than the courses promise, because the effective signal is the expensive one, and expensive signals are built slowly. In the short run you can only clean up and make legible what you already have.
Do I need a personal website? You need one if it hosts something inspectable. A site that repeats your CV adds no signal.
Is it true recruiters check social media? The most-cited figures come from self-reported surveys with weak methodology. The only large-scale field experiment finds small average effects, with subgroup exceptions. Which makes cleaning up sensible, not building a shop window.
Does personal branding matter more for freelancers? Yes, because a freelancer doesn’t have an employer’s signal behind them and has to substitute for it. I’ve collected the rest in freelancing: how to start.
Sources
- Gorbatov, S., Khapova, S. N., & Lysova, E. I. (2018). Personal Branding: Interdisciplinary Systematic Review and Research Agenda. Frontiers in Psychology, 9, 2238. The review documenting the weakness of the empirical base.
- Spence, M. (1973). Job Market Signaling. Quarterly Journal of Economics, 87(3), 355-374. The signalling model and the differential cost condition.
- Connelly, B. L., Certo, S. T., Ireland, R. D., & Reutzel, C. R. (2011). Signaling Theory: A Review and Assessment. Journal of Management, 37(1), 39-67. The state of the art on signalling theory in organisations.
- Acquisti, A., & Fong, C. (2020). An Experiment in Hiring Discrimination via Online Social Networks. Management Science, 66(3), 1005-1024. The field experiment on roughly 4,000 employers.
- Scopelliti, I., Loewenstein, G., & Vosgerau, J. (2015). You Call It “Self-Exuberance”; I Call It “Bragging”: Miscalibrated Predictions of Emotional Responses to Self-Promotion. Psychological Science, 26(6), 903-914. The gap between prediction and actual reaction.
- Burks, S. V., Cowgill, B., Hoffman, M., & Housman, M. (2015). The Value of Hiring through Employee Referrals. Quarterly Journal of Economics, 130(2), 805-839. The measured value of internal referrals.
- Bertrand, M., & Mullainathan, S. (2004). Are Emily and Greg More Employable than Lakisha and Jamal? American Economic Review, 94(4), 991-1013. The classic experiment on how signals in applications are read.
- Granovetter, M. S. (1973). The Strength of Weak Ties. American Journal of Sociology, 78(6), 1360-1380. Why reputation travels through peripheral contacts.
- Fombrun, C., & Shanley, M. (1990). What’s in a Name? Reputation Building and Corporate Strategy. Academy of Management Journal, 33(2), 233-258. Reputation as the outcome of observable signals, in the organisational version.
This article was written following the process described in how we write. If you find a figure reported incorrectly, write to me: I’ll correct it and flag it at the bottom of the article.



