Personalizing offers for customers means proposing to each one the argument that actually convinces them, instead of a single lever used on everyone. Facing an undecided customer, the most common lever is still the discount: it is the fastest one to pull, but it often answers a question the customer never asked.
Two customers undecided about the same product can be hesitating for opposite reasons. The first is waiting for a lower price, the second wants confirmation that the product is really the right choice for them, regardless of cost. Treating both the same way, with the same offer, convinces one and leaves the other indifferent.
The sections below cover why the discount stays the default answer, what actually separates the purchase motivations of two different customers, and how Portrait reads that motivation in the language of conversations the company already has with each one.
Why the discount is the wrong answer
Facing an undecided customer, the most common move for a sales team is to offer a discount. It is the easiest lever to pull: it does not require understanding what the customer is actually weighing, and it works more or less the same way on any deal. It also produces a visible result quickly, which is why it becomes the default answer even when price is not the real reason for the hesitation.
The cost shows up first in margin. Every percentage point of discount granted comes straight out of the margin on that sale, and if the lever gets used on every uncertain deal, the effect adds up across the whole quarter's revenue. A company that discounts out of habit ends up selling more units at a lower margin, without ever checking whether the discount actually convinced someone who would otherwise have gone elsewhere.
There is also a less visible effect, tied to what the customer learns from the experience. If every hesitation gets rewarded with a price cut, the customer learns that raising a doubt is worth doing, because a discount tends to follow. The next negotiation starts from a higher expectation, and the lever loses strength precisely with the customers the company has already discounted before.
A discount stays a useful tool when price really is the obstacle. The problem is using it as the first answer to any hesitation, without knowing whether that doubt is about price or something else: the guarantee, the time to get set up, the feeling of making the right choice compared with the alternatives. Working out which motivation drives that specific customer, before picking the offer, keeps margin from being spent on a lever that does not answer the real problem.
What actually drives a purchase
Two customers buying the same product, at the same moment, are often not buying the same thing. One customer picks a subscription plan because, among the options on the table, they go straight for the lowest monthly fee: they compare the numbers and stop at whichever one looks lowest. Another customer picks the same plan because they want something that makes them feel different from their colleagues, or lets them show they picked something out of the ordinary. The product bought is identical, the reason behind the purchase is not.
This difference decides which offer convinces and which falls flat. For the first customer, an offer built around exclusivity or personalization says little: what they are after is a lower price, and every other lever is noise next to that decision. For the second customer, an offer built only around price communicates the opposite of what motivates them: if the product is cheap, it becomes less useful for setting them apart from everyone else. Sending both the same offer convinces one and loses the other, even though on paper they bought the same product on the same day.
Facing this variety, a sales team usually works with a single lever applied across the whole customer base: the discount, because it is easy to use without having to tell cases apart. The result is an offer that speaks well to whoever decides on price and stays neutral for whoever decides on something else. Purchase motivations are not random: they show up in how a customer talks about a product, in what they justify about their own choice, in what makes them hesitate, well before the price negotiation even starts. The practical problem is that they are rarely asked about directly: a customer reveals them while solving a real problem, in a support ticket or a quote request, without anyone ever asking outright why they would buy.
Reading motivation in language
Portrait measures purchase motivation through the aggregated traits, a group of 38 indicators valued from 0 to 1 that come from combining several base traits. Two families relate directly to purchasing behavior: Purchase-, which describes how a customer behaves when buying, for instance whether they tend to buy on impulse or driven by emotion, and PurchaseMotivation-, which describes why they buy: what they are after in a product beyond its stated function.
The traits in the PurchaseMotivation- family separate different motivations. PurchaseMotivation-SeekBelonging marks a customer looking for brands that reinforce a sense of belonging to a social group. PurchaseMotivation-SeekStatusDisplay marks a customer focused on showing others an image of success, leaning toward expensive brands. PurchaseMotivation-SeekSelfEnhance marks a customer mostly after the chance to stand out and express their own identity through what they choose. Three different motivations, each one responding better to a different kind of offer.
Like Portrait's other aggregated traits, these are predefined correlation matrices, developed in collaboration with the Massachusetts Institute of Technology (MIT), that combine base traits measured on the customer's language into a single value ready to use.
A high value on a single PurchaseMotivation- trait tells only part of the story: read alongside Purchase-HighSpending or Purchase-Impulsivebuyer, it helps estimate how much the customer is willing to spend on that offer and how quickly they might decide to buy.
The value is calculated on the same texts a company already collects in the course of the relationship: an email explaining what the customer is after, a chat asking for detail on a specific feature, a note the sales team leaves after a call. No dedicated survey is needed, and the score updates as new text comes in. It stays in step with the latest conversation available.
Building the right offer for each customer
Building an offer this way starts from the same texts described above, the ones a company already collects in the relationship with the customer. No need to collect anything new or run a customer survey: the profile is calculated on conversations that already exist.
Once the profile is calculated, three steps guide the choice of offer:
- read the value of the relevant trait, for example
PurchaseMotivation-SeekSelfEnhance, inside the correlation matrix that generates it - compare the value against other profiled customers, to tell a strongly present motivation from one that is only a light signal
- pick which aspect of the offer to lead with based on that motivation, keeping the product fixed and changing only what gets said about it
For the same product, what changes is the order of the information and the language used to present it. For a customer with a high PurchaseMotivation-SeekSelfEnhance, it pays to lead with what makes the product different from the obvious choice: a personalization option, a limited edition, a detail other customers do not have. For a customer with a high PurchaseMotivation-SeekBelonging, it pays to lead with who else already uses the product and how it fits into a shared context. A discount, when it is genuinely needed, stays available for whoever decides on price: the difference is knowing who to offer it to, instead of handing it out to everyone by default.
The trait points to which argument to open with. The rest, available budget, purchase history, any stock constraints, stays information only the person selling can weigh case by case, alongside what the profile says about the customer in front of them.
To turn these traits on inside the platform, the page on using aggregated traits covers the steps.