Every person on your list already has an association with you. You did not choose it, nobody sat in a meeting and agreed it, and it is doing more to determine your results than any subject line you will write this year.
For a great many emails, the association is some version of: this lot always want me to buy something.
Worth being clear about what that is and what it is not. It is not an opinion somebody formed about your last campaign, and it is not something they could articulate if you asked them. It is a learned response, built quietly over months, and it now sits in front of every email you send, colouring the thing before it has been read. Your beautifully written value-led newsletter arrives and gets processed by somebody who already knows, before they open it, what you are probably after.
The useful part is that associations are learned, which means they are built by a mechanism, and a mechanism can be pointed somewhere deliberately. So this post is about how that mechanism works, what it means for the first few weeks after somebody joins your list, and how to end up associated with a subject that matters to your audience rather than with a permanent request for money.
The relevant science is evaluative conditioning, reviewed by De Houwer, Thomas and Baeyens across twenty five years of research in Psychological Bulletin in 2001, and confirmed by a meta-analysis from Hofmann and colleagues in 2010.
The finding is straightforward and slightly unnerving. Take a stimulus somebody feels nothing about, pair it repeatedly with something they do feel about, and the feeling transfers to the neutral thing. The effect has been demonstrated with images, words, tastes, smells and sounds, across a wide variety of procedures, and it does not require the person to consciously notice that any pairing has taken place.
Now map that onto an inbox. Your sender name is the neutral stimulus, and it starts out entirely neutral, because on the day somebody subscribes it means almost nothing. What it gets paired with, over and over, is whatever feeling arrives alongside it. And after twenty or thirty pairings your name stops being neutral, because it now carries whatever it has been arriving with.
Every email is a pairing event, so it helps to be blunt about what the feeling attached to each one really is:
An ask. Mild obligation, or mild irritation, depending on the day and how many of them there have been.
A discount for something they were not thinking about. A small nudge of pressure, plus the faint sense of being sold to.
A deadline that was not real. Distrust, and it compounds, because the next deadline gets discounted before it is read.
Something useful they did not expect. Surprise and a small debt, which is the pairing you want.
A question answered before they had to ask it. Relief, which is a stronger feeling than most marketing ever generates.
Nothing much at all. The neutral pairing, which builds nothing, and which I will come back to because it is the most common outcome of all.
So what your name comes to mean is not what you have said about yourself in your copy. It is the accumulated feeling of everything that has arrived carrying it, and the accumulation happens below the level anybody is paying attention to.
The second piece of the picture comes from how memory is organised. Collins and Loftus set out spreading activation in 1975, describing memory as a network of linked concepts where activating one concept spreads activation to the things connected to it. Keller applied the idea to brands in the Journal of Marketing in 1993, modelling brand knowledge as an associative network with the brand as a node and everything linked to it as its associations.
Which changes the goal. Being remembered is not the achievement, because a brand node with nothing much attached to it is a name that never gets activated by anything. What matters is what you are linked to, and how many different things can lead somebody back to you.
The practical version of that question: when your audience thinks about the problem you solve, does anything in that thought connect to you? Not when they think about buying, which is rare, but when they think about the problem, which is constant.
The one everybody fears, and the one described at the top of this post. They always want something. Every email is a campaign, every campaign is an ask, and the pattern has been learned so thoroughly that even the most useful email gets read as a preamble to a pitch.
What it costs you is not only unsubscribes, and this is the part that gets missed. A negative association degrades the performance of good work you have already paid for, because the email that would have worked on a neutral audience arrives at an audience that has already decided what it is.
Almost nobody worries about this one and almost everybody has it. No strong feeling in either direction, no particular subject attached to your name, nothing for activation to spread from. You are a sender they vaguely recognise and cannot connect to anything they care about.
Neutral feels safe because nothing is going wrong, and it is the least recoverable of the three positions, because there is no cue anywhere in your audience's mental network that leads back to you. You are not disliked. You are simply never the thing that comes to mind, which produces exactly the same revenue as being disliked while feeling considerably more comfortable.
They help me with something. Not they sell something I sometimes buy, which is a product association and only activates in the rare moments somebody is in market, but they know about a subject I care about, which activates every single time the subject comes up.
That difference is the whole strategy, so it gets its own section.
If your association is with your product, you can only be retrieved when somebody is thinking about buying your product, and they are almost never thinking about that. If your association is with the subject around your product, you get retrieved constantly, including at the exact moment the need starts forming.
Two examples, deliberately from industries that have nothing in common, because the principle does not care what you sell.
A hair care brand
The product association is: they sell shampoo. Activated perhaps six times a year, when a bottle runs out, in competition with everything else on the shelf and whatever is on offer.
The subject association is: they understand my hair. Curly hair that never behaves in humidity, a scalp that flares up in winter, postpartum shedding, the damage from colouring, what to do about it. Activated every time somebody looks in a mirror and is annoyed, which is not six times a year, and which is nearly always the moment before a purchase eventually happens.
A construction firm
The product association is: they build things. Activated when a tender goes out, which for any given client is once every few years if you are lucky.
The subject association is: they understand where projects go wrong. Planning risk, building regulations, why programmes slip, what a realistic contingency looks like, how to spot a contractor who has underpriced. Activated in every board meeting where a project is discussed, in every conversation about a scheme that is running late, and long before anybody writes a specification.
Same mechanism in both cases, and it works for a charity, a SaaS platform, a wholesaler or a freelance bookkeeper. You are choosing to be the answer to a question people ask often, rather than the answer to a question they ask rarely.
The subject sits at the intersection of three things, and the intersection is usually narrower than people expect:
What you are properly expert in. Not what you would like to be known for, what you could talk about for an hour without preparing.
What your audience worries about frequently. Frequency is the whole point, because a subject that comes up once a year is barely better than the product association.
What leads to buying you, eventually and honourably. The subject should sit upstream of your product rather than being unrelated to it, otherwise you build a lovely association that never converts into anything.
Asch published his work on forming impressions of personality in 1946, showing that the order information arrives in changes the impression it produces. The same list of traits, presented in a different sequence, generated a different overall impression, because whatever came first set the frame that everything after it was interpreted through.
Your first few emails are that frame. Whatever they establish becomes the lens for the next two years of sending, and it is considerably easier to set a frame than to change one.
So look at what most programmes do in the fortnight after somebody subscribes. A welcome email that mostly talks about the company. A discount, because the signup was a discount. Two product emails. A sale. The frame that gets set is commercial, it gets set within about ten days, and every useful email sent in the following year has to fight it.
Which is why I keep arguing that welcome flows should be orientation flows. The job of the first few sends is not to convert somebody who just arrived, it is to establish what your name is going to mean.
Oliver set out expectation-disconfirmation in the Journal of Marketing Research in 1980, modelling satisfaction not as an absolute judgement but as a comparison between what was expected and what arrived.
Applied to email, the promise you made at signup becomes the denominator for everything that follows. A weekly email full of insight is a delight to somebody who was told to expect occasional updates, and a disappointment to somebody who was promised daily deals. Identical email, opposite association, and the difference was set by a sentence on a form that nobody has looked at in three years.
Worth going and reading your own signup form before you spend another month optimising anything downstream of it.
Kahneman, Fredrickson, Schreiber and Redelmeier published a study in 1993 in which participants held a hand in painfully cold water, then repeated the trial with an extra period during which the water was warmed slightly, so the second trial contained strictly more discomfort in total. Asked which they would rather repeat, a majority chose the longer one, because it ended better.
Redelmeier and Kahneman found the same pattern in a clinical setting in 1996, where patients' retrospective evaluations of an uncomfortable procedure were predicted by the worst moment and the final moment rather than by how long the discomfort lasted overall. The finding is now generally called the peak-end rule.
So nobody is keeping a running average of your emails. They are holding a strongest moment and a most recent moment, and those two things are doing the work of summarising you.
Three practical consequences follow from that.
You need peaks, not consistency alone. A programme of reliably decent emails with no standout moment is filed as unremarkable, because there is no peak for the memory to hang on.
The last email before a gap matters more than the ones before it. If your final send before a quiet period is an ask, the ask becomes the summary of you until the next one arrives.
Your unsubscribe experience is an ending. The last thing somebody experiences of your brand gets weighted heavily, and plenty of people who leave a list still buy, still recommend, and still talk. A gracious exit is not generosity, it is the final pairing.
Name the association you want, in one sentence. When I think about ______, I think of them. Write it down, agree it with whoever writes, and treat it as a constraint rather than a slogan.
Audit what you are currently pairing yourself with. Take your last ten sends and write the dominant feeling next to each one, not the topic. Most people find eight asks, one announcement and one thing that was useful.
Make the early sends carry the subject. The frame gets set in the first fortnight, so whatever you want to own has to be visible before the first serious ask arrives.
Hold your sender identity completely still. One sender name, one voice, one promise. An association cannot form around something that keeps changing shape, and every sub-brand you invent starts the conditioning from zero.
Vary the format, repeat the subject. Repetition of the subject is what builds the link; repetition of the format is what makes people stop noticing. Those are two different kinds of repeating and most programmes get them the wrong way round.
Fix the ratio. Every ask is a pairing with obligation. If asks outnumber gives, the conditioning is running against you no matter how good the writing is.
Build in peaks on purpose. A few times a year, send something markedly better than your baseline. Something that costs you real effort, gives away something you could have charged for, or says a thing your competitors would not.
End every sequence well. Especially before a quiet period, and especially at unsubscribe.
Association does not show up in a click rate, so you have to go and look for it in places most reporting ignores.
The language people use when they reply. Read your replies for the words people use about you, because that is the association in their own phrasing rather than yours.
One survey question, asked once a year. What do you come to us for? Free text, no options, and the answers will tell you what you have built rather than what you intended.
What people say when they refer you. Ask a customer how they described you to somebody else. The sentence they use is your association, said out loud.
Whether unprompted contact happens. People replying with questions about your subject, rather than about an order, is the clearest evidence the link has formed.
You are conditioning your audience whether or not you are doing it on purpose. Every send is a pairing, the pairings accumulate, and after a year or two your name means something specific to several thousand people who could not tell you why.
The question worth answering is not whether you are building an association, because you already are. It is whether the one you are building is the one you would have chosen, and whether anybody in your business has ever written it down.
Most programmes are quietly building either mild irritation or complete neutrality, and both of those are expensive in ways that never appear on a report. The alternative takes longer, costs less than most people assume, and compounds, because trust and influence are what an association becomes once it has had enough time to settle.
Research referenced: De Houwer, Thomas & Baeyens (2001) and Hofmann et al. (2010) on evaluative conditioning, Psychological Bulletin · Collins & Loftus (1975) on spreading activation, Psychological Review · Keller (1993) on customer-based brand equity, Journal of Marketing · Asch (1946) on forming impressions of personality · Oliver (1980) on expectation-disconfirmation, Journal of Marketing Research · Kahneman, Fredrickson, Schreiber & Redelmeier (1993) and Redelmeier & Kahneman (1996) on the peak-end rule.