Skip to content
MakerGain
← Back to cases

He Launched Two AI Products at the Same Time: The Media Chased the Buzz, the Money Went to the Other

Edited by MakerGainPublished: Article updated:

Danny Postma launched Deep Agency and HeadshotPro in the same period: the former was chased by media coverage, while the latter was the one that saw payments first. What carries real weight is the action after that contrast: at around six months after launch, he had gone three consecutive weeks without building new features, and the growing number of competitors made him anxious; during the same period he posted the first job opening for a machine learning engineer and began moving from a solo operator to a small team; as for “refund if it doesn't look like you” and non-expiring team credits, those are terms the company states on its pricing page.

For example, someone who wants to change their LinkedIn headshot: they pull out their phone, take a few selfies and upload them; two hours later they receive thirty professional headshots, pay 29 dollars, settled in one go. pricing page[1] HeadshotPro compresses the entire chain of “booking a photographer, gathering everyone, scheduling, and waiting for retouching” into uploading a few selfies. [2]

Its backstory is a little interesting. Founder Danny Postma recalled in an interview with Unite.AI that he launched two products during the same period—Deep Agency and HeadshotPro. The former received a great deal of media coverage but did not convert into sales; demand for the latter appeared quickly, so he shifted more of his energy to HeadshotPro. [3]

What the media chased and where the money went were not the same direction.

After Uploading, Only One Action Remains

The personal plans come in three tiers and are one-time purchases. Basic is $29 for 30 images, delivered in about 2 hours; the more expensive tiers shorten delivery time and increase the number of images, and Executive also includes 4K and landscape/portrait versions. Each tier includes 1 free redo within 60 days and is covered by a 30-day Realism Guarantee: “refund if it doesn't look like you.” pricing page[1]

Image: source HeadshotPro actual image source[4], retrieved 2026-11-10, original interface retained.

The failure this category fears most is not ugliness; it is an acquaintance recognizing at a glance that “this is not him.” So the refund clause is not a marketing add-on; it addresses the last bit of hesitation before the user presses the pay button: what if the person who comes out doesn't look like me—what happens to this 29 dollars? The official site makes “looking like you” a core product requirement—professional headshots need to be polished, and they also need to be you. [2]

Behind the scenes, to make this work, the back end has more piled up than the marketing slogan suggests. Danny said in the interview that, beyond basic generation, the company had deployed dozens of open-source and in-house models to filter out unacceptable uploads and NSFW results, erase AI artifacts on faces, and had already put one year into developing its own custom model. [3]

In the same interview, he mentioned that at the time the product required uploading 15–17 photos; this was the empirical number he had found maximized quality, and the team was looking for a way to upload fewer photos without sacrificing quality. The team page captured on November 14, 2026 describes another team workflow: employees upload 1–3 selfies with their phones. The page does not say when this number took effect or which plans it covers, so it cannot be treated as a revision record of the same workflow; what can be established is that he was already looking for ways to reduce uploads at that time.

In Team Orders, the Buyer Is Not the Employee but the Person Who Manages Employees

This line is another offering HeadshotPro built specifically for organizational customers. It takes on the trouble of “who organizes it.”

Administrators first lock in a background and clothing pool and invite employees by email or link; employees upload 1–3 selfies from their phones and choose one of the approved styles; each person receives 40 high-resolution headshots and 1 brand-template headshot, with the page claiming they are returned within 30 minutes; administrators track progress and bulk-download from a dashboard. [5]

What is solved here is not only “looking good in the photo” but also “who organizes it.” Traditional shoots require gathering everyone and scheduling, then finishing in one session; when a new employee joins, they often have to wait for the next shoot or pay again. HeadshotPro sells credits as “one full shoot per person”; each credit includes 40 headshots and 1 free retry, the credits do not expire, and they can be saved for new employees. [1]

The pricing page lists 100 credits at $25.35 per shoot (35% off), for a total of $2,535; the discount rises from 20% to 50% with cumulative purchase volume. The page explicitly states no subscription and no automatic renewal; additional charges consist of an optional retouching credit pack and SSO billed at the identity provider's cost. [1]

This point is worth pausing on. HeadshotPro is a one-time payment; Photo AI, which also does personal AI photos, says on its site “plans from $9/mo billed yearly”—an annually billed subscription. The two products have different product forms, so they cannot be directly compared on price. But HeadshotPro removes the hassle of “forgetting to cancel” from the process. [6]

To sell this into companies, images alone are not enough. The team page says it acts as a data processor and the customer is the data controller, and it provides a DPA; uploaded photos are automatically deleted 30 days after generation, data is encrypted, and photos are not used to train AI; it also claims SOC 2 Type II certification. Administrators can enforce two-factor authentication, and IT can also log in with company accounts (SAML 2.0 SSO and JIT provisioning). These are not selling points; they are tickets into the procurement process. [5]

The site itself also names the alternative: traditional shoots usually cost $100–$300 per person, and each reshoot or new employee requires paying again; the page claims that, counting photographer fees, travel, and lost work hours, teams save an average of $209 or more per person. This is the seller's own calculation, broken down in one of its cost articles. [5]

Those Three Weeks in the Middle

The period at around six months after launch may be the part of this article that least resembles a startup story. In an interview with The Bootstrapped Founder, he said he had gone three consecutive weeks without building new features and had lost the motivation to keep developing, and that the growing number of competitors made him anxious; at the time, his wife helped with customer service for 2–3 hours a day, and he was still a solo operator, but he had already posted the first job opening for a machine learning engineer and was starting to think about a small team. [7]

Earlier, the servers often went down while he was sleeping; later the system became stable enough that customer service could wait. While away for two weeks, he once simply turned on automatic refunds so customers with problems could get their money themselves without waiting for him to reply. [7]

In that Unite.AI interview, he said that going from a solo operator to a team was a major challenge, that he was afraid of hiring, and that at first he tried to automate everything as much as possible; among the three options—selling, remaining independent, and building a team—he chose to build a team, starting with contractors and part-time people who already had other clients. [3]

He had been down this road once before. While working on Stock AI, he changed direction because quality fell short of expectations and he was also worried about legal risks from large companies such as Getty; he then built ProfilePicture.AI in about 30 hours, it blew up on Twitter, generated significant sales within a week, and even ranked near the top in Google's people headshot search—after the buzz died down, sales fell with it. [3]

The Money and the Unclear Parts

The official site's About page claims that since launch it has generated 17,000,000 professional headshots, completed 200,000+ shoots, and brought in $10,000,000+ in cumulative revenue. The page does not give a cutoff date for the statistics. Revenue for the current period, gross profit, generation and redo costs, customer service and team expenses are all undisclosed, and the share of revenue from teams is also not disclosed, so profit remains unclear, and the average order value or number of customers cannot be inferred from these three numbers. [2]

What is clear is the contrast itself: Deep Agency's media coverage was real, and HeadshotPro's payments were also real. Danny saw demand for HeadshotPro appear quickly and shifted more of his energy to it.

“Refund if it doesn't look like you” catches the risk that the photos are unusable; “credits do not expire” catches changes in team hiring and scheduling. A 29-dollar headshot package sells the former; teams stockpiling credits per person-shoot buy the latter.

Continue reading

AI publishingAI writingAI imagesAI video
Do Not Start From a Blank Page: CaptureFlow Catches What You Say; the Signature Is Still Yours

CaptureFlow catches expression from voice, links, video, or files and hands it to Content Agent to draft; only after the user revises and explicitly approves does it publish to LinkedIn, and responsibility remains with the person. Chris Koronowski runs the company in Poznań, Poland, charging in three tiers—Pro, Business, and Agency—with the latter two priced by seat; a case study says InStream has 100 paid seats, while TrustMRR separately shows 8 subscriptions, and the two cannot be combined into a customer count.

AI productivityAI writingAI images
A Polish-language AI with no registration, dismantling the registration barrier down to the payment page

The Polish-language chat service Czat.ai lowers the barrier to use by not requiring registration or English and by accepting BLIK payment; founder Jaskólski says the subscription, account, and prompt barriers were deliberately removed. The website shows 28 roles and file and image delivery, while the terms make the role definitions and refund boundaries clear. TrustMRR records show that it has paid subscriptions, but it does not disclose profit or renewal retention, so profitability cannot be judged from them.

AI companionsAI imagesAI audio
One reply counts as one, a photo as eight: Embraces.AI breaks companionship into messages and sells it

Embraces.AI prices companionship by message: one reply counts as 1 message, and one Picture this image counts as 8. Companions have cross-session memory, send multiple bubbles in a row, and initiate contact; users can edit memories and export chats. Before signing up, there are 40 free messages; the three paid tiers start at $12, and $5 tops up 250 messages. TrustMRR recorded on 2026-11-14 all-time revenue of $48,193 and 426 active subscriptions.

AI imagesAI videoAI audio
59 Million Views but Almost No Revenue: After Publishing the Entire Project, Overseas Signups Increased to 6.5x on the Day

Zohar Vanunu used his self-built Kolbo.AI to make a “biblical phone imagery” AI series, which drew 59 million views and more than 200,000 followers in 30 days, but by his own account it brought in almost no revenue; after publishing the complete production tutorial, overseas signups increased to 6.5x on the day, and in three weeks 3,285 people accounted for 78% of new signups. The product puts more than 100 models, along with chat, image, video, music, and voice, into the same project and charges by credits.