Author: Josiah Cole

Categories
Business Web Development

19 Ways Business Owners Torpedo Their Own Website Project (2026 Update)

Last updated: October 17, 2026

Disclosure: some links in this post are affiliate links, marked as sponsored. If you buy through them I may earn a commission at no extra cost to you. I only recommend hosts I have used on real client sites.

I first wrote this list in 2010, when I had about 14 years of client work behind me. It is now closer to 30, and the list holds up better than I’d like. The tools changed (nobody asks for a Flash intro anymore), but the ways a project goes sideways are mostly the same.

I love my clients. They pay my bills, they build real businesses online, and they mostly listen to me. This post is not about any of them in particular, I promise. It is a field guide to the habits that make a good project worse, so you can skip them.

Paint cans and brushes
Let the people you hired do the painting.

Design and feedback mistakes

  1. Design it yourself. If you hired a competent designer, let them turn your ideas and feedback into something you love. That is what you paid for.
  2. Blindly follow a friend who “does graphics”. A trusted second opinion is good. Letting one person’s taste override the plan is not.
  3. Go silent, then send everything at once. Weeks of no contact followed by a frantic evening of notes buries your team. Set aside a regular time each week instead.
  4. Don’t communicate at all. Most web professionals want client input, even messy input. Silence is the hardest thing to design around.
  5. Say you like a design you hate, then nitpick it. Be honest early. Bending a design you never liked into shape will not give you a site anyone is proud of.
  6. Hold your ideas until the first proof. Share examples, competitors, and must-haves at the start. Feedback is cheapest before anything is drawn.

Content mistakes

  1. Ask your developer to write your content. Most developers are not copywriters. Plan for content from day one, and ask your team to recommend a writer if you need one.
  2. Reuse the copy from your old site. There is a reason you are redesigning. Old copy usually carries the old site’s problems with it.
  3. Skip the photographer and the copywriter. Photos sell the product and copy helps customers find you. Friends and family rarely substitute for a professional.
Child with a camera
Your nephew’s camera skills are not a content strategy.

The 2026 version of mistake 7: “AI will write it”

AI tools are useful for outlines, first drafts, and alt text. They are not a substitute for knowing what your business does better than anyone else. Generic AI copy reads like everyone else’s generic AI copy, and search engines and customers both notice. Use it to get started, then rewrite it in your own voice with your own details.

Money and schedule mistakes

  1. Pay the deposit late. Work starts when the project is funded. A late deposit pushes everything back.
  2. Pay the deposit, then disappear. Projects that stall for a year come back with stale plans, new requirements, and higher rates. Every day the site is not live is a day it is not working for you.
  3. Hand over one giant feedback document. Small, regular rounds of feedback are easier to act on than a 30-page binder.
  4. Set a hard deadline, then wing it. Deadlines are fine. Pair them with a realistic plan, and agree to adjust when things change.
  5. Make your own “better version” of the design. Mockups in MS Paint (or Canva) rarely help. Describe what you want in words and let the designer do the design.
  6. Invent your own site structure. Some clients want to reinvent how navigation works. Customers want a site they can find their way around without thinking. Stick with what works.
  7. Keep the old logo no matter what. You don’t need a full rebrand, but letting your designer freshen a dated logo often pays off.
  8. Ask for features from 2007. Splash pages, autoplaying music, and pop-up “launch site” buttons. See my webmaster commandments for the full list.
Sale sign
Cheap now often means expensive later.
  1. Underpay. If a quote sounds too good to be true, it usually is. A site built for a third of a fair price tends to cost the difference again when you pay someone to fix it. Look for a partner who can support you for years.
  2. Overpay. Overpaying bothers good technology people almost as much as underpaying. Get a few competing quotes, and ask each provider why the others cost more or less.

Mistake 20 (new for 2026): picking hosting last

The project ends, the site is ready, and then someone asks where it will live. Hosting chosen in a hurry usually means the cheapest shared plan, and that choice affects speed, security, backups, and how much your developer can help later.

My short guide:

Building a site yourself? WordPress.com’s AI builder gets you a working first draft, with hosting included.

Try the WordPress.com AI builder

FAQ

What is the most common reason website projects fail?

In my experience it is content. The design and code are usually ready long before the text and photos are. Plan who writes your content at the start of the project.

How often should I give feedback to my web designer?

Small amounts on a regular schedule, such as once a week. It keeps the project moving and avoids a pile of conflicting notes at the end.

Can I use AI to write my website content?

You can use it for outlines and first drafts, but rewrite the result with your own details, prices, and voice. Unedited AI copy tends to sound like every other site.

How many quotes should I get for a new website?

Two or three is usually enough. Ask each provider to explain what is included, so you are comparing the same scope and not just the price.

When should I choose web hosting for a new site?

Early, ideally before development starts. Your host affects staging, backups, speed, and security, and your developer may have strong preferences based on how they work.

Related posts

Categories
Crazy Ideas

I’m Stuck: A Random Design Idea Generator (and What I’d Build Today)

Last updated: October 17, 2026

Every designer hits a wall. In 2010 my answer was a tiny web app called “I’m Stuck”. You clicked a button and it gave you one random design suggestion: change the color palette, try a serif, crop tighter, remove something. Most suggestions were useless for the project in front of you. Every so often one knocked something loose.

The original demo lived on josefresco.com and is no longer online, so this post is the record of the idea and where I’d take it now.

The original idea

The concept was simple: a list of short, specific design prompts and a button that picked one at random. No accounts, no settings. Just a nudge.

It works for the same reason a walk works. When you’re stuck, you’re usually stuck on one path. A random constraint forces you onto a different one.

Ideas for expanding it

These were my plans in 2010, and most still make sense:

  • Ratings. Let people vote suggestions up or down so the useful ones surface.
  • User submissions. Let other designers add their own prompts.
  • Filters. Narrow suggestions to typography, color, layout, or copy.
  • Community. Show what people made after following a suggestion.
  • Mobile. In 2010 this meant a mobile site. Today it would simply be a responsive page you can add to your home screen.
  • Beyond design. Random pick-up lines, economic policy, sitcom plots. The format works for any creative block. Some of these would work better than others.

The 2026 update: AI and an older idea

Two things have changed how I think about this.

First, I later learned the idea was not new. Brian Eno and Peter Schmidt published Oblique Strategies in 1975, a deck of cards with prompts like “Use an old idea” for musicians and artists who were stuck. If you like the concept of “I’m Stuck”, that deck is the original and still the best version of it.

Second, AI chat tools can now generate endless prompts on demand. That is useful, but it misses part of the point. A short, fixed list of human-written prompts works because each one is blunt and a little strange. An AI asked for “design ideas” tends to give you the safe, average answer, which is the opposite of what you need when you’re stuck.

If I rebuilt it today, I’d keep a curated list of prompts and use AI only for an optional second step: “Here’s my project and the prompt I got. Show me three ways to apply it.”

Try it without an app

You don’t need software. Write 30 short prompts on index cards, shuffle them, and draw one when you’re stuck. A few to start with:

  • Remove one element and see if anyone misses it.
  • Make the most important thing twice as big.
  • Use only one typeface.
  • Design it for a phone first.
  • Swap your main color for its opposite.
  • Write the headline before you touch the layout.

FAQ

Is the I’m Stuck web app still available?

No. The original demo on josefresco.com is no longer online. The prompts in this post and a deck like Oblique Strategies cover the same idea.

What are Oblique Strategies?

A deck of cards created by Brian Eno and Peter Schmidt in 1975. Each card has a short prompt meant to help artists and musicians break out of a creative block.

Can AI help with designer’s block?

Yes, as a second step. Pick a blunt, specific prompt first, then ask an AI tool for ways to apply it to your project. Asking AI for open-ended ideas tends to produce safe, average results.

How do I get unstuck on a design project?

Add a constraint. Limit yourself to one typeface, design for a phone first, or remove one element. Constraints push you off the path you’re stuck on.

Related posts

Categories
Web Design

A Webmaster’s 19 Commandments: My 2007 Rant, Revisited for 2026

Last updated: October 17, 2026

Disclosure: some links in this post are affiliate links, marked as sponsored. If you buy through them I may earn a commission at no extra cost to you. I only recommend hosts I have used on real client sites.

I wrote this list in 2007, angry about splash pages and websites that resized my browser. It became my most popular post ever, read by more than a quarter of a million people. A lot of the technology in it is gone. Flash is dead, nobody worries about dial-up, and Firefox is no longer the browser you have to remind people about.

What surprised me when I reread it is how many of the rules still hold once you swap in the modern equivalent. Below is each commandment, what I meant in 2007, and my verdict for 2026.

The 19 commandments, then and now

1. Do not resize the user’s browser window

2007: A bit of JavaScript that resized the window was the mark of a spam site or a designer who “also does websites”. 2026: Browsers mostly block this now. The modern version is hijacking scrolling or the back button. Same rule: the visitor controls the browser, not you.

2. Do not make visitors “launch” your real site in a pop-up

2007: Splash page, click “Enter”, new window. You lose. 2026: Splash pages are rare, but full-screen newsletter pop-ups that fire before anyone reads a word are the same mistake. Deliver your message first.

3. Don’t make people pick a “high bandwidth” or “low bandwidth” version

2007: Asking visitors to choose their connection speed. 2026: Responsive design and modern image formats do this for you. One site should work on every screen and connection.

4. Don’t build the whole site in Flash

2007: Flash was a great tool for animation and video, and a terrible way to build an entire site. 2026: Flash is gone. The modern trap is a heavy JavaScript app for what should be a simple brochure site. If search engines and screen readers struggle with it, rethink it.

5. Don’t reinvent navigation

2007: Clever, mysterious menus that make people hunt for the links. 2026: Still true. Put navigation where people expect it and label it in plain words.

6. Print designers, stop “doing websites on the side”

2007: This one got me in trouble. Print and web are different crafts. 2026: Softer verdict. Plenty of great designers work in both. The point stands, though: the web has its own rules for speed, accessibility, and responsiveness, so hire someone who knows them.

7. Put real text on your home page

2007: Text baked into images can’t be read by search engines. 2026: Still true, and now it matters for accessibility and AI search too. Use real text, and real alt text on images.

8. Make sure it works in Firefox

2007: Too many sites only worked in Internet Explorer. 2026: Test in Chrome, Safari, Firefox, and Edge, and above all on phones. Most of your traffic is probably mobile.

9. No blinking text, no FrontPage, no pop-ups, no scrolling text

2007: A grab bag of things that had already been beaten to death. 2026: Blinking text and FrontPage are history. Autoplaying video and stacked cookie, chat, and newsletter overlays are the new versions.

10. No music

2007: Background music that starts on page load. 2026: Browsers now block most autoplaying sound. Don’t fight them.

11. Use text navigation

2007: Image-based menus were common and hard to read for search engines. 2026: Still good advice. Icons need text labels too.

12. Have a site map

2007: A page listing every section of the site. 2026: You need an XML sitemap for search engines, which WordPress creates for you at /wp-sitemap.xml. A human-readable one is optional.

13. Add search

2007: Any site with more than a few pages needs a search box. 2026: Still true. WordPress includes one, and better search plugins exist if the built-in results are weak.

14. Load time still matters

2007: Over half of American internet users were still on slow connections. 2026: Connections are faster, but pages are much heavier, and Google measures speed with Core Web Vitals. My guide to speeding up WordPress starts with your host, because that is where most slow sites begin.

15. Don’t hide your message

2007: A home page is a billboard. Give people a message and a path. 2026: Still the most important rule on this list. One clear headline and one clear next step.

16. Give visitors predetermined paths

2007: Guide people toward what you want them to do. 2026: Same thing, now called a conversion funnel.

17. Respect bandwidth with video

2007: Don’t force a large video on people who didn’t ask for it. 2026: Host video on a service built for it, lazy-load embeds, and never autoplay with sound.

18. Don’t make people mouse over things to find out what they are

2007: If users have to hover to discover what a link does, find another line of work. 2026: Even more important now, because phones have no hover at all.

19. New technology is not a reason by itself

2007: Just because something is new, or you just discovered it, doesn’t mean the site needs it. 2026: Today that technology is usually AI. Use it where it helps visitors, not because it’s on the feature list.

Bonus rules from readers

After the post spread on Reddit, readers added a few rules I still agree with:

  • Don’t link to a PDF without saying it’s a PDF.
  • Don’t use scripts that stop visitors from leaving with the back button. Locking customers in your store doesn’t make them buy.
  • Proofread. “You’re” and “your” are not the same word.
  • Don’t write in random CAPITAL LETTERS for emphasis.

Building a site in 2026 without breaking the rules

The easiest way to follow most of these commandments today is to start from a modern platform that handles mobile layouts, sitemaps, and speed for you. If you’re building a small site yourself, WordPress.com and its AI website builder give you a sensible starting point. Read my notes on what to fix before you launch an AI-built site first.

Starting a new site? WordPress.com’s AI builder sets up a responsive site with hosting included.

Try the WordPress.com AI builder

FAQ

Are Flash websites still possible in 2026?

No. Adobe ended Flash Player support at the end of 2020 and every major browser has removed it. Animations and video now use HTML5, CSS, and JavaScript.

Do I still need a site map?

You need an XML sitemap so search engines can find your pages. WordPress generates one automatically. A visible site map page for people is optional.

What is the modern version of a splash page?

A full-screen pop-up or overlay that appears before the visitor sees any content. It has the same effect: people leave before they learn what you offer.

Which browsers should I test my website in?

Chrome, Safari, Firefox, and Edge, on both desktop and mobile. Mobile Safari and Chrome on Android matter most for most sites.

Is this advice still relevant for WordPress sites?

Most of it, yes. WordPress and modern themes handle the technical items like sitemaps and responsive layouts, but clear messaging, simple navigation, and speed are still up to you.

Related posts

Categories
Announcements

Hello World: The First Post on JosiahCole.com (December 2005)

Last updated: October 17, 2026

Disclosure: some links in this post are affiliate links, marked as sponsored. If you buy through them I may earn a commission at no extra cost to you. I only recommend services I have used myself.

The original post

Here it is, exactly as I wrote it on Christmas Eve 2005:

This is the first post on Josiah Cole.com.

That’s it. One sentence. The URL still carries the original title, “these things are sprouting like weeds”, which is how blogs felt in 2005. Most of the blogs from that year are gone. This one is still here.

Twenty years of posts

If you just landed here, this is a quick tour of the archive, one or two posts per era. Each one has been updated for 2026.

I also post link dumps of bookmarks I want to share instead of hoard.

Thinking about starting your own blog? WordPress.com’s AI website builder drafts a starter site from a short description, so you can skip the blank page.

Try the WordPress.com AI builder

FAQ

When did JosiahCole.com start?

The first post went up on December 24, 2005. The blog has been running, on and off, ever since.

What does this blog cover?

Mostly WordPress, web hosting and running a web business, plus gadgets, games, AI tools and whatever else I can’t stop thinking about.

Where should a new reader start?

The posts listed under “Twenty years of posts” above are a good tour. They have all been updated for 2026.

Related posts