Your Kid Finished Their First STEM Kit - Now What?

The window after a finished kit is short. A child who has just built something works at their peak confidence for about two weeks - and if the next project doesn’t arrive in that window, the interest usually cools rather than compounds. Most parents buy the wrong second kit not because they choose badly, but because they choose by price or age instead of by the thing their kid actually enjoyed.

There are only four things a child can fall in love with in a first build: the assembly, the code, the device’s purpose, or the finished object’s status. Work out which one it was and the next purchase makes itself. This guide covers how to read the signals, four progression paths with specific next steps and prices, and the two upgrade decisions (soldering, and text-based code) that determine how far the path goes.

Step One: Diagnose What They Actually Liked

Watch what happened during the build rather than asking what they thought of it. Kids are terrible at self-reporting and excellent at revealing.

The Builder

• What you saw: Wanted to do every screw themselves, complained when helped, immediately looked for something else to take apart
• What they loved: The assembly
• What to buy next: A harder build: more parts, longer, eventually soldering

The Programmer

• What you saw: Finished the build quickly, then spent hours changing the code or making their own game
• What they loved: The code
• What to buy next: A device with a deeper software ceiling, then text-based programming

The Practical User

• What you saw: Kept using the finished device for its actual function - playing, messaging, wearing, driving
• What they loved: The purpose
• What to buy next: A different domain, not a harder kit: music, wearables, robotics, comms

The Collector

• What you saw: Showed everyone, took photos, wanted it on a shelf
• What they loved: The object
• What to buy next: Something impressive and collectible: licensed replicas, big builds

Two honest signals in the other direction: if they abandoned it halfway and you finished it, the kit was too hard or too long - go sideways, not up. And if they finished it dutifully with no visible enthusiasm, the domain was wrong. A kid uninterested in robots may be very interested in music hardware. Don’t conclude they’re “not a STEM kid” from one box.

Path 1: They Loved Building

This child wants more parts, more steps, and eventually a real tool in their hand.

Next step (same difficulty, more content): a two-device kit. Chatter 2.0 ($149) is a pair of encrypted communicators, roughly two hours for both - twice the assembly of a typical single kit, and the second unit goes faster than the first, which is a confidence experience worth engineering deliberately.

Then: the soldering decision. This is the biggest single jump available, and the reason to make it is that soldering opens the entire hobby - every Arduino project, every repair, every own-design circuit. The CircuitMess NASA Mars Perseverance Rover ($349) is the serious version: roughly 20 hours of assembly with servo and DC motors, a robotic arm, a rotating camera tower, radio control and a Mission Control module. It’s a multi-week project, not an afternoon.

Before that, get the bench right - a soldering project on a dining table with no fume control is how good projects become bad memories. The Complete Workshop Bundle ($59) covers the tools, heat-resistant mat and goggles together; our workbench setup guide covers ventilation, lighting and the safety rules that actually matter.

Path 2: They Loved the Code

This child treated assembly as the tax on getting to the software. Give them a device with a deeper ceiling, then change notation.

Next step: a programmable robot. Wheelson 2.0 ($169) is a self-driving car with a camera and object recognition - line following, obstacle detection, autonomous driving - programmable in CircuitBlocks and then Python and C++. Robots reward better code visibly, which is exactly what a code-motivated kid needs.

Then: the text-code transition, which is the other big jump. Don’t make it a subject change - rewrite a program they already understand, on a device they already own. Our guide to moving from block coding to real Python sets out the read-then-modify stage most self-taught paths skip.

Good value option: the Coding Bundle 2.0 ($269) packages ByteBoi 2.0, Bit 2.0 and Codee 2.0 - three programmable devices, which keeps a software-focused kid supplied for months and works well when two siblings both want one.

Path 3: They Loved What the Device Does

This is the most commonly misread signal. The child isn’t asking for a harder kit - they’re telling you which domain they care about. Change the subject, keep the difficulty.

• Music. Synthia 2.0 ($149) is a buildable synth, sampler, sequencer and drum machine; Jay-D 2.0 ($169) is a DJ mixer with real DACs, amplifiers and a crossfader. The Music Bundle 2.0 ($279) pairs them. Why it works: sound is instantly gratifying and endlessly tweakable - see music tech for kids.
• Wearables. Clockstar 2.0 ($99) is a build-and-code smartwatch. The NASA Artemis Watch 2.0 ($129) ships assembled and programmable. Why it works: the finished device gets worn all day, so the project stays present - see wearable tech for kids.
• Vehicles. The Rimac Nevera ($199) is a 1:28 scale, no-soldering, programmable replica of the Croatian hypercar, with its own build-it-yourself controller - see how electric cars work.
• Communication and privacy. Chatter 2.0 ($149) leads into cryptography and radio - see encryption for kids.
• Virtual pets and care mechanics. Codee 2.0 ($89) builds a programmable virtual pet with a real-time clock and six mini-games; the Beep Boop digital pets are $39 each or $195 for all four.

Path 4: They Loved Owning It

Status and collection are legitimate motivators, and dismissing them wastes a real lever. The trick is to make the collectible thing require work.

• Licensed and limited builds. The Rimac Nevera ($199) and the Butter Bot ($309) - an AI-powered robot inspired by Rick and Morty with extra modules - are display-shelf objects that had to be assembled first.
• Expansions rather than new kits. If a child owns a Bit 2.0, the Bit Expansions are $8.99 each, each unlocking a game and 14 tradable cards; the Complete Wacky Expansion ($59.99) covers all nine, and the Wacky Collector’s Bundle ($125) includes the console plus all nine expansions. Cheap, frequent, collectible - good for maintaining interest between big builds.
• Bundles as milestones. The Bestsellers Bundle ($339), Epic Gaming Bundle ($619) and Ultimate STEM Bundle ($1,199) work as a birthday or graduation event rather than a casual purchase.

The Two Real Upgrade Decisions

Everything above is domain choice. Only two decisions actually change what a child can do.

Decision 1: Soldering, yes or no. There is no official minimum age - readiness is fine motor control, following instructions, and care around hot objects, which many kids show around 9-11 with an adult present. Saying yes opens the entire electronics hobby; saying no still leaves years of screwdriver-assembly kits. Our guides on how to teach kids to solder and the best soldering kits for kids cover the how.

Decision 2: Text code, yes or no. Blocks are genuinely good pedagogy - in one controlled study, students learning with blocks outperformed those learning in text - but ambitious kids eventually stop respecting them. The signals for readiness: they find dragging blocks slow, they read the generated code unprompted, and they want a capability the palette doesn’t offer.

A kid who has crossed both lines by 13 has functionally unlimited runway: any Arduino or Raspberry Pi project on the internet becomes available to them.

Keeping the Gap Warm (Free)

You don’t need to buy something every fortnight. Five ways to hold momentum between kits:

1. Re-programme the device they own. A new game, a new watch face, a new drum pattern. The hardware is already paid for and the software ceiling is usually nowhere near reached.
2. Set a constraint challenge. “Make the robot follow a line without using the camera.” Constraints produce more learning than new features.
3. Get it out of the bedroom. A school show-and-tell, a demo for grandparents, a video. Audience is a renewable motivator.
4. Break something on purpose and fix it. Deliberate debugging practice, which is the actual skill.
5. Teach someone younger. Explaining a build to a sibling consolidates it better than any worksheet.

If the gap runs long, a subscription smooths the cadence: the STEM Box tiers run from $139 (Apprentice) to $679 (Hero) and $1,199 (Legend), delivering kits on a schedule instead of you deciding each time. Whether that’s worth it depends on your household - our comparison of STEM subscription boxes versus one-time kits makes the case both ways.

When to Stop Buying Kits Entirely

The honest endpoint. Once a teenager can solder, read text code, and use a multimeter, kits become a slower and more expensive route than the alternative: a $30 microcontroller, a parts assortment, and their own idea. Signs you’ve arrived - they want a part rather than a project, they’ve started modifying kits instead of finishing them, or they’re browsing GitHub for someone else’s build.

At that point your job changes from buying kits to funding components and protecting bench time. That’s the outcome the whole progression is aiming at, and it’s what our piece on what engineers wish they’d learned as kids describes.

Frequently Asked Questions

What should my kid build after their first electronics kit?

Match the next kit to whichever part of the first one they loved. If it was the assembly, go longer and harder - eventually a soldering project. If it was the code, get a programmable robot with a deeper software ceiling. If it was what the device did, change domain rather than difficulty: music, wearables, vehicles, communications. If it was owning the finished object, pick something collectible or licensed that still has to be built.

How soon should the next kit arrive?

Within a couple of weeks of finishing, while confidence is high. That doesn’t have to mean a purchase - reprogramming the device they already own, or a constraint challenge on the existing build, holds the momentum just as well and costs nothing.

Should the second kit be harder or just different?

Different, unless the child clearly enjoyed the difficulty itself. Parents routinely over-escalate, buy something too long, finish it themselves, and conclude the child lost interest. If a first build was abandoned halfway, go sideways to a shorter kit in a domain they care about.

When is a kid ready to start soldering?

There’s no official minimum age; no regulator publishes one. Readiness is about fine motor control, reliably following instructions, and demonstrated care around hot objects - often around 9-11 with direct adult supervision. Set the workspace up first: ventilation or fume extraction, impact-rated goggles, a heat-resistant mat, and hand-washing after every session.

Are bundles better value than buying kits one at a time?

Per unit, yes - bundles like the Coding Bundle 2.0 ($239 for three consoles) or the Music Bundle 2.0 ($225 for two) cost less than the separate kits. But the pacing matters more than the maths: three kits arriving at once can produce one built device and two boxes in a cupboard. Bundles work best for households with multiple kids, for gift occasions, or for a child with a demonstrated appetite for volume.

What if my child just isn’t interested after the first kit?

Distinguish domain from activity. A child bored by a robot may be genuinely engaged by a synth, a smartwatch or an encrypted radio - the engineering content is comparable and the motivation is completely different. If several domains fall flat, park it for six months. Interest at 9 and interest at 11 are frequently different children.

The Bottom Line

The second kit matters more than the first, because it decides whether one afternoon becomes a habit. Read which of the four things your kid actually loved, buy for that rather than for age or price, and make the two real upgrades - soldering and text code - deliberately rather than accidentally. The full range at CircuitMess covers every one of those four paths, which makes it straightforward to change domain without starting over.