Korea’s first large-scale, longitudinal study of the consumer ‘Say-Do Gap’
tracks 29,182 consumers over two years, showing that stated purchase intent
remains a powerful demand signal - intenders went on to buy at nearly than
double the rate of non-intenders.
The Say-Do Gap Series
I: Intention vs. Purchase
within Two Years
- First domestic study to quantify the consumer intention-to-purchase
conversion rate, based on an unprecedented large-scale, longitudinal dataset
of individually matched respondents.
- In 2023, 52% of tracked consumers said they intended to buy a vehicle
within two years.
- By July 2025, only 25% of those intenders - one in four - had actually
purchased a vehicle.
- Among consumers with no purchase plan, 13% went on to buy a vehicle
anyway.
- Of all actual buyers, roughly two-thirds had originally intended to buy
and one-third had not.
- Among planned buyers, only 44% purchased their originally preferred brand,
and just 21% purchased their exact preferred model.
SEOUL, South Korea - April 23, 2026 - ConsumerInsight, a leading automotive market research firm, today released
new findings on the consumer `Say-Do Gap` - the difference between what
consumers say they intend to do and what they actually do - based on
large-scale automotive survey data matched at the individual respondent
level.
The analysis asks a simple question: among consumers who said in 2023 that
they planned to buy a vehicle within two years, how many had actually done
so by July 2025? The answer: about one in four. At the same time, even among
consumers who said they had no plan to buy, about one in eight purchased a
vehicle anyway. It is the first time this intention-to-behavior gap has been
measured in Korea using individually matched, longitudinal data at this
scale.
One in Four Intenders Followed Through
In the 2023 survey, 52% of matched respondents said they intended to buy a
vehicle within two years [Table 1]. By the 2025 follow-up,
13% of all matched respondents were consumers who had expressed that
intention and then actually purchased. Converted to the base of planned
buyers, the actual-purchase conversion rate was 25% - meaning only one in
four who said they would buy actually did so within the period.
The gap also showed up on the other side of the market. Respondents with no
two-year purchase plan made up 48% of the matched sample. Among them, 6% of
the total sample - 13% within that group - purchased a vehicle anyway,
roughly half the conversion rate of the intention group.
Overall, 19% of matched respondents purchased a vehicle by 2025. Of those
buyers, about two-thirds had originally said they intended to buy, and
one-third had not - underscoring that stated intention, while informative,
captures only part of who ultimately enters the market.
[Table 1] Comparison of Automotive Purchase Intention and Actual
Behavior (The Say-Do Gap)
Based on continuous respondents (N = 29,182)
| Categroy |
% of Total Respondents |
| With Purchase Intention |
52 |
| └Actually Purchased |
13 |
| └Did Not Purchase |
39 |
| Without Purchase Intention |
48 |
| └Actually Purchased |
6 |
| └Did Not Purchase |
42 |
| Total Actual Buyers |
19 |
A Meaningful Signal, Despite the Gap
ConsumerInsight said the results point to a meaningful distance between
stated plans and actual behavior: only 25% of intenders followed through,
and 13% of non-intenders bought anyway. At the same time, the purchase rate
among intenders was nearly double that of non-intenders, confirming that
purchase intention remains a valuable - if imperfect - demand signal for
automakers and marketers.
The Gap Extends to Brand and Model Choice
The Say-Do Gap was not limited to whether consumers bought a vehicle - it
also appeared in what they bought. Among all actual buyers, 31% purchased
the brand they had originally preferred, and 14% purchased their exact
preferred model [Table 2]. On the base of purchase
intenders alone, the equivalent rates were 44% and 21% - meaning only about
one in five intenders ended up with both their preferred brand and preferred
model. ConsumerInsight noted that real-world constraints such as available
budget, new model launches, delivery timing, promotions, and taxes can
significantly reshape a consumer’s final choice.
[Table 2]
Profile Analysis of Actual Car Buyers
Based on actual car buyers (N = 5,495)
| Buyer Segment |
% of Total Buyers |
Notes |
| Planned Buyers |
70 |
Had purchase intention prior to buying
|
|
└Purchased Preferred Brand
|
31 |
Of the 31 points, 14 also involved the exact preferred model
|
|
└Purchased Alternative Brand
|
39 |
|
| Unplanned Buyers |
30 |
Had no purchase intention
|
From Theory to Data: Measuring the Say-Do Gap at Unprecedented Scale
The Say-Do Gap has long been discussed in academic and industry circles,
largely on a theoretical basis or through small-scale, cross-sectional
studies. ConsumerInsight’s analysis is significant for quantifying the
phenomenon using individually matched, longitudinal data from roughly 29,182
consumers tracked over two years - a scale with few, if any, precedents
worldwide.
Building on this release, ConsumerInsight plans to continue publishing
multi-angle analyses of the Say-Do Gap - examining which consumer segments
show the widest gaps, which vehicle categories are most affected, and how
economic conditions or new model launches shape the shift from intention to
purchase. The goal is to help automakers and marketers design and execute
more precise, evidence-based strategies.
Notes
- Say-Do Gap: In this release, the term refers to the difference between a
respondent’s stated 2023 two-year vehicle purchase intention and actual
vehicle purchase behavior observed by the July 2025 follow-up.
(ConsumerInsight Annual Automobile Syndicated Study, 2023-2025 matched
analysis)
- Survey base: Figures are based on 29,182 respondents who answered both
the July 2023 and July 2025 surveys. The survey population consists of
licensed drivers who own a vehicle or plan to purchase one, so results may
differ from a general-consumer survey. (ConsumerInsight Annual Automobile
Syndicated Study)
- Percentage bases: The 52% purchase-intention share is among all matched
respondents. The 25% conversion rate is among purchase intenders only. The
13% figure for intenders who purchased is the share of all matched
respondents who both intended and bought. (ConsumerInsight Annual
Automobile Syndicated Study)
- Brand/model basis: Preferred-brand and preferred-model purchase rates
are reported as 31% and 14% on the actual-buyer base (N = 5,495), and as
44% and 21% on the planned-buyer base. (ConsumerInsight Annual Automobile
Syndicated Study)
Methodology
ConsumerInsight used data from its Annual Automobile Syndicated Study,
conducted every July since 2001 with an annual sample of roughly 100,000
automotive consumers (95,696 respondents in the 2025 wave). This analysis
matched 29,182 respondents who participated in both the 2023 and 2025
waves and compared each respondent’s 2023 stated intention to purchase
within two years with their actual vehicle purchase behavior as of the
July 2025 survey.
- Population: Licensed drivers who own a vehicle or plan to purchase one
within two years
- Sampling frame: ConsumerInsight’s Invight Panel and quota-based samples
from major Korean portal-site panels
- Sampling method: Gender- and age-based quota sampling
- Survey mode: Email and mobile survey
- Survey content: Usage & Attitude (U&A) and Consumer Experienced Quality
(CEQ)
- Survey period: July 2023 and July 2025
About ConsumerInsight
ConsumerInsight is a leading South Korean market research firm
specializing in automotive consumer data and analytics. Since 2001, it has
conducted the Annual Automobile Syndicated Study, surveying approximately
100,000 automotive consumers every July - one of the largest continuous
automotive consumer studies in the world. Drawing on more than two decades
of longitudinal data, ConsumerInsight provides in-depth insight into
consumer behavior, market trends, and industry benchmarks.
For more information, please visit
https://www.consumerinsight.co.kr/eng/
Copyright ⓒ ConsumerInsight. All rights reserved.
This material may be
quoted or cited by media for news purposes. Any other form of reproduction,
distribution, or secondary use requires prior written consent.