Josh Sprague on Orange Mud, Product Testing, and AI

Josh Sprague Built Orange Mud by Refusing to Confuse a Working Prototype With a Finished Product

Josh Sprague's first hydration pack was made from a gun holster, a ratchet strap, an old waist pack, and a hotel sewing kit. It looked terrible. It also answered the only question that mattered that night: could a bottle ride high on a runner's back without bouncing all over the place?

The answer was yes. That did not mean the product was done.

Orange Mud says its products are designed, tested, and refined by athletes who use them. Image and source: Orange Mud.

Josh walked back into the house and told his wife that the experiment had two possible outcomes. They would either build a company or end up with two backpacks that had cost them $25,000 each. That rough prototype eventually became the starting point for Orange Mud, the endurance-gear company he founded in 2012. Fourteen years after the first experiment, he came on Venture Step to explain what happened after the clever origin story.

That is where this conversation gets useful. A prototype can prove that an idea is real. It cannot prove that the product will remain comfortable after three hours, fit bodies unlike the founder's, survive repeated abuse, arrive from manufacturing with the right details, or earn enough trust that someone will depend on it during a race.

The first successful test created a harder question

Josh had spent years racing marathons, triathlons, mountain bikes, and adventure events. He knew the irritation that started the company because he had lived with it. Waist packs squeezed and moved. Handheld bottles occupied a hand he did not want to use. Traditional backpacks trapped heat or placed weight where the body moved most.

His founder advantage was not simply passion. It was access to a problem, a community, and a demanding test environment.

He could take a prototype on a long run and wait for it to reveal what a quick jog concealed. In the interview, he said roughly three hours was the minimum useful run for understanding certain endurance-product failures. Pressure points, movement, rubbing, and fit problems could remain invisible for the first hour.

Three hours is not a universal testing rule. It is a reminder that every product has a point where the demo ends and the customer's real experience begins. A running pack has to survive distance, sweat, different bodies, nutrition stops, fatigue, and the constant motion of a person who no longer wants to think about their gear. Other products have their own equivalent.

The test is not whether a product works once. The test is whether it keeps working where the customer would lose trust.

The last ten percent was not one final task

Josh described getting a design to roughly 90 percent as the easy part. The remaining work was where the hidden systems collided. A pack could be stable but create a pressure point after three hours. A waist belt could make sense in principle but take years of intermittent work before it deserved to ship. One pack worked in short tests, failed after longer use, sat untouched for months, and returned as a different product.

That history matters because “the last ten percent” can sound like polishing. In Josh’s examples, it included fit, reach, body variation, material behavior, manufacturing, and whether the product became noticeable at the exact moment the customer needed it to disappear.

He also described the rare opposite case: one idea that arrived almost fully formed and worked in the first prototype. The exception made the normal pattern easier to see. Most product development was not a straight path from inspiration to production. It was a repeated sequence of use, failure, distance, return, and redesign.

The discipline is not perfectionism. A founder can always imagine one more change. The useful question is whether the unresolved problem changes safety, function, durability, comfort, manufacturability, or the trust behind the product’s promise.

Seven Clay came from the same pattern

Orange Mud was not Josh's only attempt to turn recurring frustration into an operating system. Seven Clay grew from his dissatisfaction with outsourced embroidery and finishing work.

From the buyer's side, a loose thread or poorly placed mark can look like a minor defect. From the brand's side, it is the finished product that reaches the customer. The outsourced shop may see one job among hundreds. The brand owner sees a promise with their name on it.

When Orange Mud moved to Texas, Josh brought more of that work under his own control. Seven Clay became a separate custom embroidery, screen-printing, and laser business. The company started with a built-in customer and a founder who understood the failure from both sides.

Dalton described that position as being customer zero. It is a powerful advantage when it creates fast feedback and honest standards. It becomes a liability when the founder assumes that personal experience represents the whole market. Josh's attention to different bodies, customer feedback, and repeated use is what keeps customer zero from becoming customer only.

AI became another tool for removing friction

The second half of the conversation moves into Josh's current AI workflows. He described using AI for product-image preparation, marketplace listings, HTML, advertising analysis, product research, and the unpleasant administrative work that can absorb a small team's time.

The value was not that one model could run the company. It was that a founder with years of operational context could describe a specific failure and review the result.

When an image tool changed a logo, the useful instruction was not “make this better.” It was to preserve the logo's proportions and change only the surrounding work. When a marketplace listing was difficult to construct, the advantage came from defining the required output, supplying the product information, and checking the result before it went live.

Josh also described failures, inconsistent model behavior, and tasks where one tool performed better than another. That matters. The lesson is not to pick a winner and outsource judgment. It is to use the tool that fits the job, keep claims and taste under review, and remember that a generated first pass is not evidence.

The standard that connects the whole episode

Orange Mud, Seven Clay, and Josh's AI use may look like separate topics. They share one operating standard.

Start with a real failure. Make the problem concrete. Build the cheapest useful test. Put the result into the environment where it can fail. Then keep the judgment close to the person who understands what failure costs.

“Good enough” becomes dangerous when it means the team has stopped learning. The opposite problem is endless polishing with no connection to the customer. Josh's method sits between them. Keep iterating while the evidence says a meaningful failure remains. Stop when the remaining work no longer changes trust, safety, use, or the promise the brand has made.

That is a much harder standard than producing a clever prototype. It is also how a rough object built with a hotel sewing kit becomes a company that is still operating more than a decade later.

About Josh Sprague

Josh Sprague is the founder of Orange Mud and Seven Clay. His current public site describes his work across consumer product development, brand operations, and manufacturing, while Orange Mud’s company history dates the first HydraQuiver prototype to 2012. An earlier company interview also records his background in endurance racing and medical-device contract manufacturing.

Those sources establish Josh’s public account and current projects. They do not independently verify every performance, savings, or milestone claim discussed in the episode. The canonical [[Josh Sprague Guest Profile]] preserves the entity record so the episode does not need to become a second biography.

Listen to E119 of Venture Step for the full conversation with Josh Sprague, including the endurance stories, Orange Mud's product-development process, the creation of Seven Clay, and the AI workflows he now uses across both businesses.

Readers interested in structured product evaluation should continue to E100 with Derek Hales and NapLab when that public page is available. E117 extends the product-development theme into a deep-tech actuator company. E120 is the better follow-up for the AI operating boundary that appears in the second half of Josh’s conversation.

Sources

This story is based on the Venture Step interview preserved in [[E119 Full Transcript]], Orange Mud's current company history, Orange Mud's earlier founder interview, Seven Clay's company history, and Josh Sprague’s public site. Operational savings and tool-performance statements remain Josh's experience, not independent benchmarks. AI assisted with source organization and drafting; the transcript and linked sources control factual claims, and publication remains subject to Dalton’s review.