AI Tools Among Technology Respondents: Dupple Data

Which paid AI tools did technology respondents report in Dupple's post-signup quiz? Of 7,225 people who selected Technology and answered the AI question between 5 August and 3 November 2026, 61.3% reported paying for at least one AI tool. These are self-selected new subscribers across Dupple newsletters, not a representative sample of all technology workers or a single newsletter's readership.

Published and updated 5 November 2026. Data read 3 November 2026.

Dupple publishes Techpresso, AIpresso, Devshot, Cyberpresso, Finpresso and MarketingShot, and sells sponsorships in them. Our data comes from a voluntary post-signup quiz and has not been independently audited; the figures describe answers to a specific question, not independently verified subscriptions or ad exposure.

Reported choices in the Technology segment

The denominator for each row is the same 7,225 Technology respondents who answered the AI question in the 90-day window. A respondent could name several tools, so the tool shares must not be added together. Source: the AI stack CSV, read 3 November 2026.

The table reproduces the Technology row of the source file. The file presents the most common choices for each segment, not a complete census of AI products. The named choices are the quiz labels used in the snapshot; a response does not prove that the person maintained an active account throughout the window or used a particular feature.

Who answered, and who is missing?

The quiz appears after someone signs up to a Dupple newsletter. The broader audience dataset contains 130,965 respondents across the 90-day window. The AI stack export records 28,509 answers to the AI question out of 130,964 respondents in its own source snapshot, and 7,225 of those answers are in the Technology segment. The one-person difference between the exports is retained; its cause is not established here. An answer to the general quiz is not automatically an answer to the AI question.

People chose whether to respond, and they selected their industry themselves. Those who answered the AI question may be more interested in AI than people who skipped it. We cannot measure that difference from the exported aggregates. A technology-industry selection is not a verified job title, an employer audit or proof that a respondent buys software for a team.

The quiz covers new subscribers across the Dupple network. It does not identify all current subscribers, deduplicate people across every possible newsletter system, or establish that the 7,225 respondents belong to Techpresso, AIpresso or any other single list. The audience and lead data page describes the source, definitions and refresh schedule. The Techpresso page and AIpresso page describe the publications separately.

What multiple selections mean

The Claude Pro and ChatGPT Plus figures are 35.6% and 31.6% of the same respondents. A person who named both contributes to both shares. The CSV does not show how many selected both, so subtracting one share from the other does not produce a count of exclusive users. The 4.0 percentage-point difference between the displayed rounded shares is not a controlled comparison of the products and should not be interpreted as proof of a market-wide preference.

Similarly, the 38.7% no-paid-tool response does not mean those people never use AI. A respondent might use a free product, a tool supplied by an employer or another arrangement not described by the aggregate export. The data does not distinguish who personally paid from every possible way a team might fund access. Ask what the questionnaire actually captures before using the numbers to define a buyer segment.

The source also records 61.3% reporting at least one paid AI tool. Those two shares describe the dataset's paid versus no-paid response categories and sum to 100% after rounding. They should not be added to individual tool shares, because the individual choices are not mutually exclusive categories.

The engineering subset is not a second audience to add

Within Technology, 2,035 respondents selected Engineering & Software Development and answered the AI question. In that subset, 65.4% reported paying for at least one AI tool; 37.5% named Claude Pro, 31.7% named ChatGPT Plus and 7.5% named GitHub Copilot. Source: the same AI stack CSV, read 3 November 2026.

The 2,035 engineering respondents are part of the 7,225 Technology respondents, not an additional group. Their self-selected role is broad and does not establish a verified developer title, a specific programming language or employer size. The engineering row's most-common-choice list differs from the Technology row; absence from a short list is not proof that nobody selected a product.

A developer-tool advertiser could use the engineering cut to frame a question about a campaign, while an AI product aimed at a broader technology audience might start with the Technology row. Neither cut proves that a specific publication or issue will reach a guaranteed number of those respondents. The software developers audience page provides separate context; keep its population and date distinct from this AI-question subset.

Use the data for a concrete buying question

Suppose an advertiser wants to explain an AI product to technology professionals. The 61.3% response supports a narrow observation: many people in this particular answering group reported paying for at least one tool. It does not say they are shopping for the advertised product or will replace an existing tool. A useful next step is to ask the publisher about the proposed publication, format and audience evidence, then prepare a message explaining the task the product helps with.

If the product is meant for engineering teams, the 2,035-person subset is more relevant than an unrelated finance or HR cut. Even then, a respondent's quiz choice is not a targetable list of verified purchasers. The Primary Ad page describes an in-issue sponsorship, while the dedicated-send page describes a separate advertiser-focused email. Ask which publication and date are available and how the proposed campaign will be reported.

The lead generation page describes a different post-signup format. The newsletter buying guide covers proposal questions, and the sponsorship ROI guide discusses defining a useful outcome. A quiz response, a delivered email, a click and a qualified inquiry are different events. The CSV cannot be used to calculate the number of buyers that a sponsorship will produce.

Product names and source dates

The source file records names as survey options, not a live product comparison. For current product descriptions, see the providers' own pages for Claude, ChatGPT, Gemini and Perplexity and GitHub Copilot, checked 5 November 2026. These pages describe products, not the population in Dupple's quiz. Their current plans or features may change after this snapshot.

A product name in the CSV is not a claim that a provider endorses Dupple or that the respondent used a particular feature. The export does not measure user satisfaction, frequency of use, outcomes or tool performance. It also does not establish whether one person's use was for work, personal projects or both. No provider's product is ranked by this page.

Read the file and refresh it carefully

The downloadable CSV has five overlapping segment rows: all AI-question respondents, Technology, Technology plus Engineering & Software Development, Finance & Banking, and Founders and C-suite across industries. Each row has its own answer count and reported choices. A person can belong to more than one row, so adding segment counts would double count some respondents.

Percentages are rounded in the source. Multiplying a rounded percentage by the row size may produce a non-integer and does not recover an exact respondent count. Use the denominator and the displayed share together, and do not create a precise count from a rounded value. The source also includes self-reported spend fields; this page focuses on choice shares rather than presenting those amounts as product prices.

The source was read on 3 November 2026; the data page says the next planned refresh is 3 December 2026. A refreshed export can change as the window and respondents change. Preserve the date, source file and question wording when comparing two snapshots, and do not interpret a difference as a change in the whole market without a comparable sampling method.

What we could not verify

The export does not provide a representative panel, verified payment records, tool overlap, frequency of use, work-versus-personal use or a publication-specific audience. The AI answer group is self-selected and smaller than the full quiz group. Dupple's quiz is not independently audited. Provider pages do not verify who answered the quiz, and none of the figures predicts an advertiser's campaign result.

FAQ

How many Technology respondents answered the AI question?

The source records 7,225 Technology respondents who answered between 5 August and 3 November 2026.

What share reported paying for at least one AI tool?

In that 7,225-person Technology group, 61.3% reported paying for at least one AI tool. Answers were self-reported and not checked against billing records.

Can the individual tool shares be added together?

No. A respondent could name several tools, so the choices can overlap. The exported aggregates do not show the size of each overlap.

Are all 7,225 people Techpresso or AIpresso readers?

The export covers new subscribers across the Dupple network and does not establish a single-newsletter breakdown for this group.

Does no paid tool mean no AI use?

No. The no-paid-tool response does not establish whether a person uses free tools or access supplied in another way.