In 2023, I thought I’d figured out a goldmine: buying used IT equipment from liquidation auctions and reselling it on eBay and to resellers. Servers that cost $10,000 new selling for $1,200-$3,000. Networking gear, storage devices, whole racks of equipment. The margins seemed insane.
I was going to get rich. For about six months, I thought it was working.
Here’s what actually happened.
The Opportunity (Or So I Thought)
I stumbled into this while buying office furniture from government auctions. One day, GSA Auctions had an entire datacenter liquidation lot: 20 servers, switches, patch panels, cables, everything. Bid was at $800. On eBay, a single server like these was selling for $1,500-$2,500.
The math was obvious:
- Buy entire lot: $800
- Separate into 20 server units
- Sell at market: $25,000-$40,000
- Profit: $24,000+
I was in. I won that auction. This was the beginning of everything I got wrong.
Lesson #1: Transportation and Handling Will Destroy You
That lot weighed 850 lbs. Servers are heavy. Everything is heavy.
What I didn’t account for:
- Pickup truck rental: $150 (I had to rent because I didn’t have a truck)
- Dolly/equipment: $40
- Two people to help move (paid buddies): $200
- Fuel: $60
- Damage in transit: One server got a dent and wouldn’t power on (cost me $500 sale)
Total transport cost: $450 + $500 (loss) = $950
My initial $800 lot now cost me $1,750 before I’d sold a single item.
What I learned: If you’re doing IT equipment arbitrage, you must have:
- A truck or access to one
- Packing materials (foam, boxes, anti-static bags): ~$200-$500
- Experience moving server-weight equipment
- Ability to lose 1-3 items to damage (budget it in)
If you’re doing this as a weekend side hustle and you have to rent a truck every time, margins evaporate.
Lesson #2: Testing and Verification Takes Time (and Money)
I thought I could just list the servers and sell them. Not quite.
What actually needed to happen:
Power testing: Every server needs to turn on and boot to BIOS. I did this with a PDU (power distribution unit) I rented.
- Rental: $50/week
- Electricity: $10/week
- Time: 2-3 hours per server
Cleaning: Server dust is nasty. Professional cleaning was too expensive, so I did it myself with compressed air and isopropyl alcohol.
- Materials: $40
- Time: 3-4 hours total
Hard drive wiping: I can’t sell equipment with someone else’s data on it. I used DBAN (free software) to wipe drives securely, but it took time.
- Time: 8-10 hours (servers sit in wipe queue overnight)
Photography: Multiple angles, serial numbers, condition shots.
- Time: 3 hours total
Research: What are these models actually selling for? What’s the market? What are specs?
- Time: 5-8 hours
Total time for 20 servers: ~40 hours
Realization: I was making ~$10-$15/hour on this lot, and that’s after sales, not before shipping.
Lesson #3: The Specs Matter (And I Didn’t Know Them)
I thought all servers were the same. They’re not.
A server with:
- 2-core processor vs. 8-core: $500-$1,000 difference in resale
- 16GB RAM vs. 64GB RAM: $300-$600 difference
- Old SCSI drives vs. modern SSDs: $200-$800 difference
- 1U rack vs. 2U rack: Different markets, different buyers
I listed one server as “1U Server, As-Is” without looking up the actual processor, RAM, and drive config. I got $400. Later, I realized that exact model with those specs sells for $1,800-$2,200 on eBay.
The server: Same condition, same specs, just listed better. My lack of knowledge cost me $1,400.
What I learned: You need to know IT hardware. If you don’t:
- Learn the specs (CPU model, generation, RAM capacity, storage type)
- Research actual comps (eBay sold listings, specialized forums, reseller pricing)
- Understand the market (who buys this? What do they pay?)
Or hire someone who does. I didn’t, and I left money on the table constantly.
Lesson #4: Shipping Is a Nightmare
Servers are fragile and expensive to ship.
Shipping cost estimates I didn’t account for:
A 1U server (approximately 18"×19"×2", weighs ~40 lbs):
- UPS Ground: $180-$280 (to most of US)
- FedEx Ground: $150-$250
- Freight: $200-$400 (if multiple units)
Now, if you’re selling a server for $1,500, shipping $200 is fine. But if you misjudged the market and it’s only selling for $900, that $200 shipping is 22% of your profit.
What I did wrong:
- Underestimated shipping costs in my pricing
- Chose premium packaging when basic would’ve worked
- Didn’t use freight options for heavier items (cost too much)
- Got surprised by dimensional weight charges
One example: Shipped a server via UPS that should’ve been freight. UPS charged $320 for dimensional weight. I’d priced the item at $1,100. Margin just became razor-thin.
What I learned: Always calculate shipping costs BEFORE you bid or buy. Your offer price needs to account for $150-$300 shipping per server.
Lesson #5: The Market Is More Competitive Than I Thought
I assumed there were tons of buyers for server hardware. There are. But so are there tons of competitors.
Who I was competing with:
Established resellers (Amazon, eBay stores) with:
- Better pricing (they buy volume, I buy 1 lot at a time)
- Better presentation (they have professional photos)
- Better reputation (100K+ reviews vs. my 30)
Specialized wholesalers (TigerDirect, CDW returned stock) selling the same gear at competitive prices
DataCenter liquidators (companies that only do this) with relationships and volume
Direct B2B channels (companies selling to other companies, bypassing eBay)
My competitive advantage: None. I was just a guy selling on eBay.
What happened: I had to undersell to compete. A server I bought for $150 that should’ve sold for $1,200 sold for $800 because someone with a store had 10 of them listed at $800. Buyer went for quantity trust, not best price.
Lesson #6: Seasonality and Cash Flow
IT equipment sales aren’t even throughout the year.
What actually sold:
- Q1 (January-March): Good volume (businesses upgrading for new year)
- Q2-Q3: Slow (people on vacation, budgets spent)
- Q4: Chaotic (rush from Black Friday, then year-end business equipment sales)
I bought a lot in June. Tried to sell in July-August. Sales were slow. Inventory sat for 6-8 weeks. Meanwhile, I’d already paid for the lot.
The cash flow problem: I’d buy for $800-$2,000, then wait 8-12 weeks to get paid as items sold. If I wanted to buy another lot, I had to use new capital. I didn’t have the cash to keep rotating inventory. So I did 2-3 lots and then ran out of money to buy more.
What I learned: You need working capital. Real arbitrage traders have $50K-$500K to keep inventory flowing. I had $5K. That’s a hobby, not a business.
Lesson #7: Returns and Chargebacks
One buyer claimed the server was “damaged” when received. I’d packed it well, but there was a small dent (from my handling, which I hadn’t noticed).
eBay ruled for the buyer. They returned it. I refunded their $1,200. I had to pay return shipping ($200). I got the server back with a dent. I couldn’t resell it at full price anymore.
Total loss: $1,400 ($1,200 refund + $200 shipping).
That single chargeback wiped out the profit from 4-5 other sales.
What I learned: You need product liability buffer. That means either:
- Pricing in 5-10% for potential chargebacks
- Getting B2B sales where you’re less exposed
- Selling to resellers (who accept as-is) instead of end consumers
Lesson #8: Some Stuff Just Doesn’t Sell
I bought a lot with 5 older servers (2010 era). They were cheap, so I figured I’d make money.
Nobody wanted them. Too old. Specs were obsolete. I listed them for months. Eventually sold 2 at deep discount, got tired of the others, and donated 3 to charity for a tax write-off.
Total loss: $600 in unsold inventory.
What I learned: Just because something is cheap doesn’t mean you can sell it. There’s a difference between “cheap” and “profitable.” Cheap is the purchase price. Profitable is the end-to-end process.
I should’ve skipped older server models entirely. Stick to equipment from the last 3-5 years with specs that match current demand.
Lesson #9: Your Time Is the Real Bottleneck
Let me do the honest time calculation for one $2,000 profit sale:
- Scouting/bidding: 2 hours
- Pickup/transportation: 3 hours
- Testing/verification: 3 hours
- Cleaning: 1 hour
- Photography: 1 hour
- Listing/research: 2 hours
- Responding to questions: 1 hour
- Packing/shipping: 2 hours
- Dealing with any issues: 1 hour
Total: 16 hours
$2,000 ÷ 16 hours = $125/hour. That’s not bad. But that’s the best case with a profitable sale and no chargebacks.
More realistic:
- Average sale: $1,200 (not $2,000)
- Include losses/chargebacks: -5% ($60 loss on average)
- Average profit: ~$1,140
- Time: Average 15 hours per sale (some sales are fast, some are slow)
- Effective rate: $76/hour
That’s okay. But it’s not “get rich.” And it requires you to actually do the work.
The Math on Where I Actually Failed
Over 6 months, I did this part-time:
- Total capital invested: $8,000
- Total items bought: ~80 items across 8 auctions
- Total revenue: ~$18,000
- Total expenses:
- Cost of goods: $8,000
- Shipping: $2,400
- Transport/logistics: $1,200
- Materials/testing: $400
- Chargebacks/losses: $800
- eBay fees: $1,800
- Total net profit: ~$3,400
That’s $3,400 profit over 6 months, working ~10 hours per week.
Annualized: ~$6,800/year, ~$13/hour effective rate (including the chargebacks and losses).
I wasn’t getting rich. I was making slightly more than minimum wage.
And that’s if everything went smoothly. In months where I had a bad chargeback or a lot didn’t sell well, I made nothing.
What I Should’ve Done Instead
Option 1: Specialize
- Pick a specific server model/era
- Learn it deeply
- Become the expert
- Buyers come to you, prices stay high
Option 2: Go B2B
- Sell to resellers/system integrators instead of eBay consumers
- Larger margins, fewer chargebacks, volume sales
- Requires relationships and phone work
Option 3: Add value
- Buy broken servers, repair them
- Higher margins on repair + resale
- Requires technical knowledge
Option 4: Skip it entirely
- Deploy that time and capital elsewhere
- Real estate, actual business, stock market
- Would’ve made more money
The Honest Conclusion
IT equipment arbitrage can work. People do make real money. But here’s the reality:
- You need significant capital ($50K+ to do it at scale)
- You need expertise (know hardware, know the market, know logistics)
- You need systems (testing, storage, shipping, quality control)
- You need time (it’s not passive, and it’s not quick money)
- You’re competing with people (resellers, liquidators, specialists) who have all of the above
As a weekend hobby with $5-10K capital? You’ll make beer money. Maybe $500-$1,500/month if you’re focused.
As a real business? You need to either:
- Do volume (hundreds of items per month)
- Specialize (one type of equipment, get really good)
- Add value (repair, refurbish, test)
- Find a niche (sell to specific types of businesses)
I did none of those. I just bought stuff, listed it, and hoped. And I made enough to justify the time, but not enough to think I was getting rich.
What I Did Next
I quit IT equipment arbitrage and went back to office furniture. The margins are thinner, but the logistics are simpler, the market is easier to understand, and there are fewer chargebacks. I make the same amount of money in half the time.
The lesson: Sometimes the less sexy opportunity is smarter than the one that seems amazing on paper.
For Anyone Tempted to Try This
Do a test. Buy one lot. Go through the entire process:
- Pick it up
- Test everything
- List it
- Sell it (might take weeks)
- Ship it
- Deal with any issues
Calculate your actual costs and time. Real numbers, not estimates.
Then decide if it’s worth continuing. If the numbers don’t blow you away, try something else.
The difference between “heard about this on the internet and it sounds cool” and “actually did it and know what it takes” is usually a lot of wasted time and money.
Don’t let it be you.