Should've asked AI to write the post (or at least the title)
Yup, the title is correct. There isn't a sobby story behind the project, really. Claude would've thought of one, but I decided to sit and write an honest post myself for once. Soo, IpsoLux project - Brand Guard.
Brand Guard by IpsoLux came as an idea when I started facing the "brand drift" problem - the Brand Guidelines are not followed to the letter and posts, sites, socials started drifting from the document I've designed. Then I found out that the tools that monitor the brand consistency are either very expensive or they are enterprise-level oriented. Asked myself "Hey, we do almost everything with AI today, it really strange that no such instrument exist for SMBs". And did the only logical thing... asked AI for help. :)
Then started coding (or at least Claude started coding). Yeah, I am a former developer turned graphic designer and brand manager. In a few weeks - the tool was 90% ready and I was getting exited to launch it. But one thing I knew from my developer days - testing is key. Nothing can break an app that well as a human clicking wherever he sees on the screen.
So, here we are - Brand Guard by IpsoLux is officially searching for beta testers as of today. It's a free 2-week beta test. If you are interested in how the app turned out and you think it might help you in any way - you are more than welcome to sign for the beta. The spots are empty for now, if they stay that way for some time - I might let my two kids just click left and right in the app and call it a day. :D
Hope you find it interesting. And sign in for the beta - it's free.
https://brandguard.ipsolux.com/
It's a legit problem I've noticed as well. Agent doesn't do a great job sticking to my brand.
I do free video feedback for early products. Let me know if this would be helpful for BrandGuard.
Thanks a lot. Yes, any feedback from the community here will be extremely helpful. How to reach you? The landing page is still gathering beta testers, but in my mind - it won't change compared to launch (in reality - it will, after I take into account all the feedback I received, of course). How to reach out?
Happy to do that for you. Schedule your recording day here and I'll send it over to you once it's ready: https://betabloom.app/clarity-consult
Hi IpsoLux,
Well done on your project so far!
Don't you think you hid everything behind a wall well to soon? Would've been great to have some Demo on the website - for example, Google as a case study. Or let users type in their URL and create a short audit of the website (even without the brand guidelines), just to get a taste of what it does (there'll be brand inconsistencies anywhere, and AI can be quite useful to instantly spot them).
Also, isn't BrandGuard already existing as a brand in the same sector (https://www.brandguard.ai/)? Just thought to raise it now to avoid legal problems down the line and most importantly brand confusion.
Another thing is what the user does with all the findings? Do they need to action each one individually and go an manually correct them? Or does the software provide the solution and maybe can implement it too where possible? It depends on your target audience as well - if it's the design team, they're likely to be happier to manually fixing it rather than a small business owner (who are unlikely to care a lot if they missed a font or if they adhere to any brand guidelines even).
It's nice that you kept the website short and to the point - you don't offer Terms of Use? I haven't joined for beta yet, but couldn't spot a Log In page listed on the website either - is that intentional?
I hope this helps! Best of luck!
Thank you for the extensive response and overview, definitely some points I will think through. Since this is my first SaaS build that gets to this level (I had some other ideas before, but never truly finished them ) - I am probably doing a lot of the things wrong.
To the points - I am aware of brandguard ai - it's an enterprise level tool, while mine is oriented to SMBs. And quite frankly - I was not aware of the name until after I was already working and have branded my tool the same way (you can imagine my cold sweat when I found out there is a similar tool with the exact same name). However, the entity behind brandguard ai was acquired by another company some time ago and as far as I'm aware - the tool is not operational and the profiles behind it are either removed or not active for a few years now. However, I have this in mind, thank you for pointing it out.
Regarding the demo - that's actually very good idea. I was thinking about a trial period or something like that, but was not sure yet how to approach it. But I like the idea of a small demo for a website and will probably implement it soon.
For the findings after each scan - the user has two options - to manually correct them wherever the issue is (social post, website, etc) or mark it as False-positive (it's possible that a drift captured by the tool is actually intentional). Once the issue is fixed - the user can mark it as Fixed. The difference between the two options is - Fixed does not account towards the overall brand score and grade, while Mark as false-positive does and actually increases the score since it basically removed an issue form the score calculation.
Each finding does come with a suggested fix - whether it's a bradn color drift or a tone of voice inconsistency. The technical part is easy - just change the color or the typography back to the ones used in the guidelines. The Tone of voice fixes however mark what the inconsistency is and suggest a rephrased version of the text that matches the branding.
I have intentionally not implemented a way the tool to autocorrect each issue - this means rights for editing in the client's socials and website, which is a bigger security risk. But your point on the case that a founder uses the tool and gets inconsistencies - it's an idea to implement a report that can be downloaded in pdf format so the founder can easily send it to the design team with notes on what needs to be updated.
You are correct about the Terms of use - I haven't implemented that, I'll note it to do it as soon as possible.
And finally - yes, the Log in page is hidden at the moment - the people who sign in for the beta will get an email once the beta starts with links to create their account and log in. The Log in and Sign up links will be available after the official launch.
Thank you very much about your feedback, definitely got some points to think about. There are still spots in the beta test, so you are more than welcome to join.