011 min
The problem Humanization answers
Picture a team that runs a budgeting app. A customer is charged twice and writes to support. The reply arrives within the hour. It quotes a policy number, uses the passive voice and is signed by a department.
The customer gets the refund they asked for. They still feel they were handled by a form. A month later a cheaper rival sends a short message that begins with an apology, and that message is the one the customer remembers.
The team writes a second version of the refund email as a short reply from one support agent, who signs it. Nothing about the refund changes. Only the sender does.
Humanization starts from this gap between being served and being spoken to. Companies reach it in many places: a support reply, an error message, a founder's post, a sales page. In each place the choice is the same. The text can come from an institution, or it can come from someone a reader could picture.
This complaint is older than social media. A group of writers described it before most companies had a public account.
021 min
Where Humanization comes from
The Cluetrain Manifesto is the usual starting point. Four authors, Rick Levine, Christopher Locke, Doc Searls and David Weinberger, wrote it. They put it online in 1999 as a list of 95 theses and published it as a book in 2000.
Their main argument was that markets are conversations between people. A company that wants to take part in a conversation has to sound like a person, because people sense the difference at once. The authors also described how businesses actually sounded at the time: most corporations could only speak in the flat tone of mission statements and brochures.
What the Cluetrain authors added
The manifesto went beyond tone. It said a company sounds human only once it lets real employees speak for it. A script written by a committee and read by a spokesperson does not meet that test, however friendly the words are.
That claim has a practical side. It explains why humanization is not a copywriting task alone. Someone inside the company has to be allowed to write, reply and sign.
032 min
How Humanization works
The idea behind the practice is simple. People are practised at judging other people. They are not practised at judging an institution. A buyer who sees a voice, a face and a reply can decide whether it seems honest, careless or kind. A buyer who sees only a logo has nothing to judge, so trust has nowhere to start.
Teams usually pull four levers.
- Voice: wording that sounds like a person talking, in short sentences and plain words.
- People: named staff, such as a founder, a support agent or an engineer who explains a decision.
- Behaviour: replying to comments and direct messages, and giving customers' own posts a place.
- Story: the problem the founders set out to solve, the failed first attempt, what changed. A brand story gives the other three something to stay consistent with.
These levers appear as ordinary tactics. Behind-the-scenes posts, team pages on a website, videos in which a leader answers one buyer's question, live question sessions, and customer photos shown beside the brand's own. Customer posts also work as social proof, since they show that other people chose the product.
The tactics differ by stage of the buying journey. Early on, educational posts and short videos explain a problem without selling. In the middle, honest comparisons and reviews help a buyer choose. Near the end, a demo or a trial lets the buyer test the claim for themselves.
Humanization also needs to stay the same across channels. A friendly website and a cold billing email tell the buyer that the friendliness was only for show, and the buyer believes the billing email.
Voice and tone are two different dials
A person has one voice and many tones. They talk to a friend one way and to a manager another way, without becoming a different person. Mailchimp's public style guide states this split: the voice stays the same, and the tone shifts with the situation.
Teams that skip the split use one cheerful tone for every message. That includes the messages sent to customers who are angry. A brand's tone of voice is the part that bends. The voice is the part that does not.
Only real people can carry it
The Cluetrain authors held that a company sounds human only once it lets real employees speak for it. Voice rules help, but they cannot replace a person who has permission to answer.
042 min
A worked case: Mailchimp's written voice
Mailchimp, a company that sells email and marketing tools, publishes the rules its writers follow. The rules show what humanization looks like when it is written down.
The guide says the company writes like an experienced and kind business partner. It asks for a conversational voice and dry humor, and it asks writers to cut jargon and hype. Customers there are often small businesses, and the guide describes them as confused by marketing terms. Plain wording is the company's answer.
The guide also tells writers to check the reader's mood before they choose a tone. A customer who has just finished a campaign needs a different message than one who is stuck and asking for help. The company's identity does not change between the two messages. Only the mood of the wording does. That is the point where brand identity and day-to-day writing meet.
The guide also shows two limits that most articles on the subject skip.
A mascot that does not talk
Mailchimp has a mascot, Freddie. The guide lets him smile and wink, but not speak, and it tells writers not to write in his voice. The brand has a character, yet the character is shown through gesture. Text always comes from the company's voice. A mascot that spoke would have needed its own rules, and the two voices could have disagreed.
When humor is forced
The guide warns that humor which does not come naturally can do more harm than none, and it says to stay straight-faced when unsure. Many teams treat a joke as proof of personality. This guide treats a missing joke as acceptable and a bad joke as a cost.
052 min
A second case: Apple's Mac and PC
Mailchimp humanized with words. Apple once did it with characters. In a long-running ad campaign, Apple personified itself as a character called Mac, played by Justin Long, and its competition as a character called PC, played by John Hodgman.
The ad pair showed a charming, quick-witted Mac next to an awkward PC. A viewer did not need to know anything about processors to feel which one they would rather talk to.
The two cases differ in one variable: who speaks. At Mailchimp, the company's writers speak as the company. In Apple's campaign, two actors speak instead, and the company does not speak itself. Giving a brand a character is a form of anthropomorphism, the habit of seeing human traits in something that is not human. It gives viewers a person to like. It also lets the company avoid speaking itself, which is why the Cluetrain authors would have counted it as a weaker form.
What a small survey found
The evidence for characters is thin. A 2014 study at Elon University surveyed 211 students about four brands from two industries. In each industry one brand used a spokescharacter and one did not. The brand with a character drew significantly more preference.
The sample was students at one university, and the brands were chosen by the researcher. The result shows a direction. It does not show how large the effect is in a real market.
062 min
How Humanization shows up in a product team
Back to the budgeting team. They have three decisions to make, and each belongs to a different person.
A founder decides whether to post under their own name. A face with a name can carry a company's story in a way an account named after the product cannot. A designer decides how the product talks at the moments when something has gone wrong: an empty screen, a failed payment, a form that rejected the input. These small texts are read by more customers than any campaign. An engineer decides how an outage is described. A write-up that says what broke, what it affected and what changes now reads like a colleague. A write-up built from passive phrases reads like a lawyer. A product manager decides who owns the support inbox, how fast a first reply must be, and whether a reply carries a person's name.
The team also has to decide how openly to talk about failure. Humanization that shows only good news does not build trust. Buyers can tell the difference between a company that admits what went wrong and one that hides it. The honest version names the problem, says what was done and says what comes next. The same openness applies to prices and to the customers the product is not good for. Telling a poor-fit buyer so is a human thing to do, and it saves both sides a refund.
071 min
How to use Humanization
The steps below depend on each other, so the order matters.
Decide who the brand would be as a person.
Write down their age range, interests and manner, based on who the buyers are and what bothers them. Every later step needs this description.
Write voice rules from that description.
Rules without a person behind them turn into a list of adjectives, such as friendly and bold, that nobody can apply. A rule such as "use active voice and avoid jargon" can be applied.
Put named people behind the voice.
This only works after step 2. Without shared rules, each person will sound different, and the brand will seem to have several personalities.
Make behaviour match the voice.
A warm tone is a promise. Fast replies, working refunds and honest outage notes keep it. This step comes last because the first three steps decide what is being promised.
Review on a schedule.
Humanization lasts only if it stays in daily work. A single campaign is soon forgotten, and the old institutional tone returns.
081 min
How to measure Humanization
There is no standard score for humanization, so the team picks measures that match each lever.
- Median time to first reply in hours, from the support inbox.
- Share of comments and direct messages that receive a reply, counted over a month.
- A one-question survey after a support contact: "Did this company treat you like a person?" on a scale of 1 to 5.
- Repeat purchases and referrals in the months after a change, compared with the months before.
The last measure cannot prove cause, because prices and seasons also move it. It is useful only next to the others.
091 min
When Humanization fails
The team assumes that a more human brand is always safer. Several limits say otherwise.
When the product fails
A humanlike brand raises expectations. A 2025 study in Humanities and Social Sciences Communications found that consumers were less satisfied with humanlike products than with plain ones when a failure occurred. The authors report that earlier work explains this by expectations: a brand that acts like a person is expected to perform like a capable one, so a failure disappoints more. A team that raises the warmth of its brand should raise the reliability of its support at the same time.
The emoji misreading
Teams often assume that humanization means emojis and slang. Guides aimed at marketers say the opposite: it takes more than these, and consistency, openness and real care count for more. A brand that writes in slang but takes a week to answer has changed its vocabulary, not its behaviour.
Too much sharing
Another assumption is that the brand should share everything. It should not. Choose stories that are relevant to customers, and keep personal details out of them.
Humanization used to cover a weak product
A friendly voice on a product that does not do what it says is deception. The test is whether the human qualities describe something real: a person who can answer, a policy that is kept.
101 min
Humanization vs. nearby concepts
Four neighbours get mixed up with humanization. The question that separates them is how much of the company they cover.
| Concept | What it is | How it differs |
|---|---|---|
| This pageHumanization | ||
| MarketingTone of voice | ||
| PsychologyAnthropomorphism | ||
| DesignPersonalization |
Personalization can feel human when it is done well, and a form letter that uses a first name does not. The difference is whether a person or a template did the work.
?4 questions
Questions people ask
Does humanization work for businesses that sell to other businesses?
Can a brand be too human?
Is humanization manipulative?
Is humanization a one-time campaign?
§6 sources
Sources
Show all 6 sourcesHide sources
Levine, Locke, Searls and Weinberger, The Cluetrain Manifesto, 95 Theses, 1999.
Cluetrain Manifesto, an encyclopedia entry on the 1999 posting and 2000 book.
Mailchimp Content Style Guide, Voice and Tone.
Calabro, K., Humanizing Brand Personalities, The Elon Journal of Undergraduate Research in Communications, Fall 2014.
Lu, Wang and Zhang, Impact of anthropomorphism on consumer emotions and satisfaction in product failure, Humanities and Social Sciences Communications, 2025.
The Social Cat, a marketing agency glossary entry on humanizing a brand.





