Running a law firm apparently wasnβt enough to keep my calendar interesting. Now Iβm checking whether our logo disappears in dark mode, rebuilding a DIY legal platform, and figuring out how to get more people using a Zoom tool we built ourselves.
Thatβs the current state of my Inventive Journey at Miller IP Law. None of these projects comes with a magical βbusiness growthβ button. Each involves finding something that could work better, making changes, and discovering three more things that need attention.
This weekβs update covers three projects: improving our website in dark mode, updating Snap Legal with newer tools and AI in mind, and highlighting our custom Zoom tool. The common thread is simple: make the experience more useful for the people we serve.
β‘ Quick Summary
Our website needs to work when visitors choose a darker display. That means checking logos, image backgrounds, text contrast, and embedded scheduling tools rather than assuming a good light mode design will carry over.
Snap Legal is being updated for startups and small businesses that want more DIY options. Weβre looking at how newer tools, including AI, can fit into that experience. This remains work in progress.
Our custom Zoom tool is available through Zoomβs app marketplace. Building it created another job: explaining what it does and why someone would use it. Meanwhile, future Inventive Journey updates will cover more of the business beyond website improvements.
π The Website Looked Fine. Then I Switched Modes.
A website can look polished on the screen used to build it and very different on someone elseβs device. During this update, I showed how our own site changes between light and dark viewing conditions.
One obvious issue was the logo. A logo designed to look good against white can lose definition against a darker background. Images can also bring along awkward white boxes, while embedded tools may look disconnected from the rest of the page.
Weβve been working on background removal, softer image edges, better contrast, and treatments that help logos remain visible. Depending on the element, that can mean a controlled background color, lighter accents, or a different image treatment.
These changes are intended to make the site clearer and more consistent. Iβm not presenting a measured conversion increase from this work. First, visitors need to see the content and use the page comfortably. Performance claims need actual data.
One technical detail matters: device dark mode, a websiteβs own dark theme, and a browser extension that forces dark colors can produce different results. Testing one does not tell you how every other version behaves.

π€ Snap Legal Is Getting Another Look
Snap Legal began as a way to give startups and small businesses more DIY legal options. Some business owners want to do more of the work themselves. Others need to make progress while working within a tight budget.
The platform was built before todayβs wave of widely used AI tools. That creates a practical question: which parts of the experience should change now that different tools are available?
Weβre updating the platform and documenting that work in a separate weekly series. The goal is to improve how people work through the process. Adding AI simply so we can put βAIβ in a headline would make a very expensive decoration.
Potential improvements need to earn their place. Could a tool help people organize information, understand a question, or prepare a more useful starting draft? Those are areas to evaluate, not a promise that every feature is already built or available.
Budget also does not tell the whole story. A useful DIY experience should help someone understand what they are doing and recognize when they need additional help. A completed form and a sound business decision are different milestones.
πΉ The Zoom Tool Has Entered the Marketing Phase
We also built a custom Zoom tool, now available through Zoomβs app marketplace. It grew out of wanting more from the video meetings we already use.
That project is a reminder that building something and helping people understand it are separate jobs. A developer can explain a feature in detail and still leave a potential user wondering how it helps during an actual meeting.
Our next task is to highlight practical uses and demonstrate the value clearly. A product walkthrough should answer a real question, such as what becomes easier during a meeting, instead of becoming a guided tour of every button ever invented.
β Common Questions & Answers
Does my business website need dark mode?
It needs to remain usable in the viewing conditions your customers actually use. Start by checking your existing website in light mode, its supported dark theme, and any browser settings you expect visitors to use. A dedicated theme is an implementation choice; clear content and working controls are the goal.
Will fixing dark mode bring in more clients?
It may remove a frustrating experience, but that is different from proving more consultations or sales. Track what matters, such as completed bookings and form submissions, before making claims about business results.
Is the updated Snap Legal platform finished?
The work described in this update is ongoing. Weβre revisiting the platform and considering how newer tools and AI can improve the DIY experience. This update should not be read as a list of completed features or a launch announcement.
Can AI replace the judgment needed for legal work?
A useful draft still needs to fit the facts, the goal, and the applicable requirements. When evaluating AI for a workflow, I want to know what someone must review and what decisions still require judgment. Confident wording is not a quality control process.
Why spend time on a Zoom tool while running a firm?
Meetings are part of how we communicate and do business. A tool that addresses an actual meeting problem can be worth exploring. The business question is whether the benefit justifies the development, support, and marketing time it takes.

πͺ Step-by-Step Guide: Improve One Customer Experience
1. Pick one job your customer needs to finish
Choose something specific, such as booking a consultation, understanding a service, or preparing information. βImprove the websiteβ is too broad. βMake it easy to find and complete the scheduler on a phoneβ gives you something you can actually inspect.
2. Test the whole path
Start where a visitor starts and finish where they should finish. Check the page, the buttons, the form fields, confirmation messages, and any embedded tools. Use different screens and viewing modes where practical. The homepage does not get to take the test for the entire business.
3. Fix the clearest obstacle first
A disappearing logo deserves attention, but an unreadable form label or broken scheduling control may be more urgent. Prioritize the issue that prevents understanding or completion. Save decorative fine-tuning for when people can do what they came to do.
4. Give new tools a specific purpose
Before adding AI or another app, write down the task it should improve. Decide what a good result looks like and how someone will check it. If the benefit takes six paragraphs to explain, the scope probably needs another pass.
5. Review the result before expanding
Check whether the change solved the original problem. Look at user feedback and available completion data. Then decide whether to improve it further, move on, or stop. A growing feature list is easy to count; a better experience takes closer inspection.
π°οΈ Historical Context: The Business Never Stays Finished
Business websites used to be described as online brochures. That description was convenient because a brochure mostly had to sit there and look respectable. Today, many websites also handle scheduling, purchases, intake, and support.
As those tasks moved online, the quality of the interaction became part of the service. A confusing scheduler is no longer just a design issue. It sits directly between a potential client and a conversation.
Different devices and viewing preferences added another layer. The page someone builds on a large screen may be visited from a phone, inside a different browser, or with display settings the designer never checked.
DIY platforms developed around a related need: helping people take steps without starting every task from a blank page. Templates, guided questions, and structured forms offered ways to organize that work.
AI brings another set of possibilities to those workflows. It also raises a familiar product question: does the new tool help someone complete the task more effectively, or does it add another layer they have to understand?
Our projects fit into that ongoing process. The website, Snap Legal, and our Zoom tool each reflect an attempt to improve an existing experience. None creates a permanent finish line. Apparently, the finish line has a subscription to new feature requests.
π Business Competition Examples
These are hypothetical comparisons, not reported competitor results.
The polished homepage versus the usable booking process
Imagine two service firms. One has impressive photography, but its scheduling tool is difficult to read on a visitorβs phone. The other has a simpler site with clear appointment options and usable controls.
The lesson is to evaluate the whole path. Visual polish has value, but the booking process must carry its share of the work.
The template folder versus the guided DIY experience
One business offers a folder of documents. Another explains what information users need, what each step accomplishes, and where additional help may be appropriate.
The second approach illustrates a direction worth evaluating for a DIY platform. Its usefulness depends on the quality of the guidance and the actual user experience, not the number of files included.
The feature list versus the meeting demonstration
Two meeting tools may advertise similar capabilities. One lists functions. The other demonstrates a specific meeting problem and shows how the tool addresses it.
That is the marketing challenge for our Zoom tool. We need to connect capabilities to situations people recognize, then let the demonstration support the pitch.

π¬ Discussion: What These Projects Are Teaching Me
The first lesson is that Iβm often too familiar with my own business to see it clearly. I know where the scheduling link lives. A first-time visitor does not. Familiarity can make a confusing experience feel perfectly reasonable.
The second is that small details can expose larger gaps in the review process. A logo that becomes hard to see is a visible problem. It also prompts another question: what else have we only checked under the conditions we personally prefer?
Third, affordability needs to be considered alongside usefulness. A lower-cost tool still has to help someone understand the work and make progress. A cheap product that leaves someone thoroughly confused has merely discounted the confusion.
Fourth, newer technology should be judged by the task it improves. With Snap Legal, the interesting question is where AI might help organize or prepare work. The existence of a new tool does not automatically make it a good fit for every step.
Fifth, building creates obligations beyond the launch. The Zoom tool needs demonstrations, explanations, support, and reasons for people to keep using it. Shipping software does not automatically ship customer understanding.
Sixth, my own attention is a limited resource. These projects all connect to the business, but they still compete for time. I need to keep asking which changes deserve attention now and which can wait.
Seventh, documenting the process can be useful even when the work is incomplete. Sharing an issue, an attempted fix, and an unresolved question gives other business owners something more practical than a polished victory lap.
Finally, this series needs room for more than website updates. Iβm considering shifting future episodes toward other parts of building a law firm and the lessons that come with them. There are only so many ways to announce that another image background has been removed.
βοΈ The Debate: Keep Improving or Get Back to Work?
Position One: Keep investing in the customer experience.
A business should keep improving the tools and interactions people rely on.
A website is often part of a potential clientβs first experience with the firm. Clear information and usable next steps deserve attention because they shape how someone encounters the service.
An older platform also deserves periodic review. When better ways to organize a workflow become available, it is reasonable to ask whether the existing experience still serves its purpose.
Internally developed tools can begin with firsthand knowledge of a recurring problem. That experience gives the builder a useful starting point, although it still needs to be tested against other usersβ needs.
From this perspective, these projects are part of operating the business. The discipline is to connect each change to a specific benefit and check whether it delivers.
Position Two: Put a limit on the improvement projects.
A business should protect its core work by limiting how much time improvement projects consume.
There is always another image to adjust, feature to consider, or workflow to rebuild. Without limits, an improvement project can become a permanent reason to postpone more pressing responsibilities.
A founder can also mistake personal interest for customer demand. Enjoying the development process is useful motivation, but it does not establish that other people need the product.
Every tool carries continuing work. Support, maintenance, documentation, and marketing belong in the decision alongside the initial build. The price of a feature includes more than getting it onto the screen.
From this perspective, a smaller scope and a clear stopping point are essential. Choose a worthwhile outcome, finish the work needed to reach it, and reassess before opening another project.
π― Key Takeaways
- Check the experience your visitors actually see. Review the whole customer path across relevant devices and viewing settings. A good result on your own screen is a starting point.
- Give AI a job before giving it a headline. Identify the task, the intended benefit, and the review process. Evaluate the output against those requirements.
- Treat usability as part of affordability. A DIY tool should help people make informed progress. Low pricing alone does not establish usefulness.
- Budget time for explanation and maintenance. A product needs more than development. Plan for demonstrations, user questions, and the work that follows release.
π§ Potential Business Hazards
1. Redesigning without a clear goal
A string of visual edits can consume time without resolving the issue visitors face. Start with an observable problem and a defined outcome. Once it is solved, resist reopening it solely because another design idea appeared.
2. Mistaking one display test for complete coverage
A siteβs supported dark theme may behave differently from a browserβs forced color changes. Review the environments you intend to support, including embedded tools, and record the limits of what you tested.
3. Adding AI without a review process
A fluent answer can hide missing information or an incorrect assumption. Before incorporating AI into a workflow, decide how outputs will be checked and how users will understand the toolβs role. A disclaimer cannot do that operational work by itself.
4. Describing plans as completed features
Public updates can blur the line between an idea, an experiment, and a released function. Use clear language about status. People should not have to reverse-engineer a launch announcement to learn what they can actually use today.
5. Taking on more maintenance than the business can support
A website, a DIY platform, and a meeting tool each create ongoing tasks. Make those obligations visible when prioritizing work. Otherwise, the next exciting build may inherit time already promised to keeping the current projects useful.

π§ Myths & Misconceptions
Myth: A website that looks good to me looks good to everyone.
Your browser, screen, and settings show one version of the experience. Check other relevant conditions and the full customer path before declaring the design finished.
Myth: Adding AI automatically makes an older platform better.
The value depends on what changes for the user. AI should be evaluated against a specific task and checked for quality. A new feature that adds confusion can move the experience backward.
Myth: DIY means nobody needs help.
People can handle some tasks themselves while still needing guidance on others. A useful platform should make its scope clear and help users recognize when their situation calls for additional assistance.
Myth: Once the product is built, the difficult part is over.
Release starts another phase of work. People need to discover the tool, understand the benefit, and get help using it. The code may be ready while the business around it still needs attention.
π Books & Podcasts Worth Exploring
The Mom Test by Rob Fitzpatrick
A useful companion for thinking about customer conversations. Before rebuilding a platform around what we assume people want, it helps to ask questions that reveal their actual experiences and problems.
The Lean Startup by Eric Ries
Relevant to the habit of testing assumptions and learning from changes. For these projects, the practical application is to define a useful outcome before expanding the scope.
The E-Myth Revisited by Michael E. Gerber
A fitting choice for an owner dividing time between client service and business development. It provides a starting point for thinking about how the business operates beyond the ownerβs individual effort.
How I Built This with Guy Raz
A podcast to explore for founder perspectives on developing and growing businesses. Hearing about decisions and setbacks can be a helpful counterweight to launch announcements that make everything look effortless.
βοΈ Legal Cases Connected to These Business Questions
These cases provide context for digital experiences and AI. They concern different legal issues and do not establish a general requirement to offer dark mode.
Robles v. Dominoβs Pizza, LLC
In 2019, the Ninth Circuit held that the ADA applied to Dominoβs website and app because they connected customers to the goods and services of its physical restaurants. The court reversed the dismissal and remanded; it did not decide that a particular dark mode design was required. cdn.ca9.uscourts.gov
The practical connection is the importance of access to online services. Dark mode appearance is only one design consideration, and changing colors does not establish accessibility compliance.
Nguyen v. Barnes & Noble Inc.
In 2014, the Ninth Circuit declined to enforce the browsewrap terms at issue where the website did not provide sufficient notice or obtain the userβs assent. A terms hyperlink, even near relevant buttons, was insufficient in the circumstances. law.justia.com
The connection to a digital platform is straightforward: putting information somewhere on a page is different from creating an effective process for presenting terms and obtaining agreement.
Thaler v. Perlmutter
In 2025, the D.C. Circuit affirmed the denial of copyright registration for a work submitted as authored by an AI system, holding that the Copyright Act requires human authorship in the first instance. courthousenews.com
The decision addressed that application. It does not mean that every work created with AI assistance lacks protection. For a business exploring AI, the role of human contribution remains an important question to evaluate.
π€ Bring Your Business Questions to the Conversation
If youβre building a startup or running a small business, you probably have your own version of this list. Some items concern the experience you provide. Others concern the tools you use, the products you are building, or the intellectual property behind them.
At Miller IP Law, I work with entrepreneurs on patent and trademark questions. A useful conversation starts with what you are building, what stage you have reached, and what decisions are coming next.
Visit lawwithmiller.com to learn more, or schedule a free consultation at strategymeeting.com. Bring the practical questions. A half-finished idea and a realistic budget are perfectly reasonable places to start.
![]()
π What Comes Next
The current work continues: improve the websiteβs dark mode experience, keep updating Snap Legal, and help more people understand our Zoom tool. Iβm also considering a change of focus for upcoming Inventive Journey episodes so the series covers more of running and growing the firm.
The lesson Iβm taking from this week is to keep asking what each project does for the person using it. Can they see the page, understand the process, or use the tool more effectively?
My to-do list will almost certainly keep growing. I would prefer that at least some of the new entries come from learning something useful, instead of discovering that our logo has decided to become invisible again.