DFIRST Reviews

DFirst is definitely the best purchase I’ve made on AppSumo, or at least one of the best. I liked that the agent is specialized on marketing projects. You’d probably fall even more in love with it if you tried it yourself. I bought Tier 5 and I’m going to upgrade to tier 6 very soon.

But I want to give feedback for improvements:

The things that need improvement are the workspace usability. It seems like it gets stuck when you want to move it around. The assistant sort of lays out the whole plan at once, and then you give feedback bit by bit. It listens to that feedback initially, but if you have a whole list of feedback that you want to go through step by step, it doesn’t really let you. It just takes the first piece of feedback, changes that, and then you basically have no choice but to approve it because it won’t let you keep writing more feedback. You have to approve it or approve it, basically, and that’s something I didn’t like.

And for longer tasks, like if I want to define a brand launch campaign, for example, there are many little steps before getting to the final part, like making assets. There are all these micro-tasks: planning, defining the persona, branding, and so on. If I disagree with the result at any step, it’s a hassle because changing one specific result affects many things that come after it, and every time you make a change, it consumes a lot of credits in the long run. I think it should let you go step-by-step, show you things gradually, and let you approve or tweak them bit by bit instead of forcing you to approve everything at once.

I need this tool for defining launch campaigns, creating the foundations and identity of a brand when it’s being established. I want it to set up marketing plans on at least an annual level, and ideally with a three-year perspective as well. It’s important that it helps define the core elements of the marketing strategy, like which channels to use, what messages to convey, and how to differentiate in the market.

I wanted to like DFIRST far more than I do.

The reason is simple. The idea behind the product is genuinely strong. A platform that combines research, scraping, workflow automation, AI agents, ad creation, whiteboards, and multi step campaign execution could be incredibly valuable for marketers, strategists, and founders. On paper, this looks like one of the more ambitious tools on AppSumo.

And to be fair, not everything has been negative.

The Tier 4 AI model access issue I raised has now been solved, which I appreciate. That was an important issue for me because it directly affected the value of the tier I paid for. So credit where it is due, that part appears to have been addressed.

There are also moments where you can see the potential of the platform. The product vision is clear. The scope is impressive. And when you look at what DFIRST is trying to become, it is easy to understand why people get excited about it.

But the actual day to day experience has been far less convincing.

The biggest problem is reliability. Too many workflows simply do not complete properly. Nodes keep loading endlessly, flows stall, and the expected output never arrives. I have already tried the obvious fixes on my side, refining prompts, shortening instructions, changing models, adjusting the setup, and testing different approaches. Yet the same problem keeps happening often enough that I cannot trust the platform for real work.

That matters because this is not a tool where one broken output is a small inconvenience. The whole value proposition depends on workflows actually moving from step to step in a predictable way. Once that breaks, the platform quickly becomes frustrating instead of useful.

Another issue is support. Response times have been too slow, and promised follow ups have not always happened when expected. When a product is this workflow dependent, support is not some secondary layer around the tool. It is part of the product experience. If the platform is unstable and support is also slow, the result is that users get blocked twice, first by the software, then by the wait.

The AI itself is also more inconsistent than I expected. In practice, it often fails to understand the full context of the task, misses important inputs, or hallucinates outputs that do not properly match the brief. That is especially frustrating when you are trying to use it for structured marketing work, because consistency matters. If I have to constantly double check whether the AI has understood the brand, the product, the audience, or even the language correctly, then a lot of the promised efficiency starts to disappear.

The visual outcomes have also been underwhelming. Too often they feel generic, off brand, inconsistent, or simply weak in quality. For a platform positioned around serious marketing execution, that is a real concern. It is not enough for outputs to exist. They need to be commercially usable. Right now, too many results still feel like drafts that would need significant correction before they are fit for use.

The whiteboard experience is another weak point. It sounds minor until you use it repeatedly, but navigation, moving around the canvas, and handling nodes can feel unnecessarily awkward. That friction adds up quickly. In a workflow tool, usability matters just as much as raw capability.

Then there is the credit consumption. DFIRST can burn through credits quite fast, especially when workflows fail, loop, or produce outputs that need to be rerun. That makes experimentation feel expensive. A platform like this should encourage exploration and iteration. Instead, it can feel like you are paying to troubleshoot.

So for me, the experience has been mixed, but not in a balanced way. The promise is high, the vision is strong, and there are clearly useful ideas inside the product. But the current execution still feels too unstable, too inconsistent, and too support heavy for me to fully recommend with confidence.

Pros:
Strong product vision
Ambitious feature set
Tier 4 model access issue was eventually resolved
Clear potential if stability improves
Can cover a broad marketing workflow in one place

Cons:
Too many workflows get stuck or fail to complete
Support is slower than it should be for a tool like this
AI often misunderstands context or hallucinates
Visual outputs are frequently weak or off brand
Whiteboard UX creates friction
Credit usage feels too heavy when things fail

Overall, this is a product I wanted to rate higher. I can see what the team is trying to build, and I still think it could become valuable. But right now it feels more like a promising platform in need of major refinement than a dependable tool I would want to rely on every day.

If the team improves workflow stability, support responsiveness after already bee waiting for weeks, AI consistency, and output quality, this could become something genuinely strong. At the moment, though, the gap between promise and delivery is still too big.

At first glance, DFIRST seems amazing, so I was excited to test out the tool. Before diving in, I watched all their YouTube videos (which isn't much tbh), and then tried using the tool on some ongoing projects we have.

Initially, the tools seems fantastic: it does an amazing job at understanding what I want, and lays out a clear path (that I can also edit!) to achieve the goals I'm looking for. The fact that it also uses multiple AI models, and always recommends the best model for the job is also great.

So, that's the good part...

Now the bad:

1) The training is VERY surface level and incomplete. This means that I a) either need to be a pro at AI agents & workflows already to make things work, or b) waste TONS of credits to figure things out on my own. I'd appreciate a comprehensive onboarding & training experience to get up to speed with DFIRST without having to waste multiple hours of my time, and hundreds (or even thousands) of credits on trying to figure things out on my own with substandard results.

2) The AI workflows/agents aren't as smart as they seem... And by this, I mean many things – just a few examples:
• I asked them to scrape a website and Instagram to understand the brand's colors, language, fonts, and content pillars. They scraped both (apparently successfully), but still used DFIRST's colors, still used English as the language (despite the particular website being in a different language), still just guessed the fonts (incorrectly), and completely missed the content pillars, formats, and visual style. And all this happened multiple times...
• The tool fails to establish a clear, logical workflow... By this, I mean that oftentimes it doesn't reference crucial information or inputs on its own, and wastes A LOT of credits by doing so. Yes, I could technically go in manually and tweak things afterwards, but a) that takes time, and b) that means I've just wasted ~100-200 credits on useless garbage.
• Its foreign language skills are not excellent. Not bad either per se, but it does make errors in languages other than English (so far, I've tried 2 other European languages). Mostly weird capitalization errors, so not that big of a deal, but still...

3) The output is just plain bad. For customer research, I've achieved better results using a single prompt in Gemini Pro alone. The social media posts created in DFIRST are downright terrible. In fact, even within a single workflow, the visual style, the tone of voice, and even the language (!!!) of some of the posts generated are completely different. Let that sink in: the same brand, the same prompt(s), the same workflow. And the 3 posts generated look like posts for 3 different brands, operating in 3 different geographical regions (and one of those completely missed the brand's product, displaying it as a supplement, when in fact, it was a diagnostics product). Not to mention that the posts I've managed to create are terrible as well: lazy, boring, and definitely no potential for engagement, virality, etc.

4) DFIRST is very credit-hungry. I bought the top tier, and already – just by playing around with a few basic workflows (no videos, and barely any images), I've spent about 20% of my monthly credits. Yes, I know there's a relatively inexpensive way to top up my credits, and I also know the tool needs to be sustainable in the long-run... But for the ~$800 I've spent on this tool, I'd be able to buy A LOT of credits, or even multiple years of unlimited subscriptions to mature, established AI tools. At this credit consumption rate, DFIRST feels almost like a ripoff.

5) And finally: while I wasn't affected personally, I see the founders launched a similar tool on AppSumo a few years ago that they've since abandoned. I don't know what happened to their old customers, but after a quick Google search (Digitalfirst Appsumo), I've found the following:
• Customers suddenly losing access to their dashboard (apparently without a warning) – Does this mean they've also lost access to all their generated assets, prompts, databases, and references?
• Customers complaining that there's no clear way to transfer their account to DFIRST, despite it being more-or-less a mature & rebranded version of the same product. – Does this mean that those customers were abandoned?
• Customers complaining about terrible customer support. – Does this mean we can expect a similar experience with DFIRST, especially once the AppSumo campaign (and refund period) are over?
Look, I'm not saying those customers are necessarily right (I have no idea), and I'm not saying us DFIRST customers are in for a bad experience (I certainly hope not) – but this is definitely a red flag that I think all Sumolings should be aware of.

All-in-all, I'll see how DFIRST evolves over the upcoming weeks. What I'm really looking for is a complete onboarding/training experience to ensure I can use the tool as intended. And I also hope they can improve the quality of the AI outputs significantly.

I paid Tier 3 for digitalfirst their first software on AppSumo and they abandoned the project just 2 years after no support or communication. They just started to dress this after launching this new platform. If you google Digitalfirst Appsumo you can still catch the reviews and see that they just recently addressed this to make new money.
About the "new" platform it looks polished and has multiple ai agents and requests running in chain to develop and research from my feeling I'd say you could send multiple requests in one of the popular LLMs to get the same results. It's convenient to use but again seeing this company abandoning their last project just to come back like this smells very fishy to me and I wouldn't bet on that they are going to honor the deal in a year or beyond.

Okay so I came across this during a pretty busy week at the agency and almost scrolled past it. Glad I didn't.

We've been piecing together a stack of separate tools for image gen, research, video - you know how it goes. It works but it's messy. DFIRST.AI basically puts most of that under one roof, and the workflow actually flows. That part surprised me.

The brand consistency thing is genuinely useful if you're managing multiple clients. You set it up once and outputs don't come back looking completely off-brand. Still needs a human eye before anything goes to a client, but the starting point is much better than I expected.

For a lifetime deal price, it's a pretty easy yes. I've paid more for tools that do less.