Data3's AI Earnings Rally Masks A Reseller Margin Recovery

Data3 surged 14% on its earnings — and the "AI" label helped sell the move. The business economics are more ordinary than the headline implies.

Data#3 (ASX: DTL) is an Australian IT reseller and services company. On August 23, its shares jumped roughly 14% to A$10.72 after reporting full-year results ending June 30, 2026, that management called a "record." Gross sales reached A$3.4 billion, up 12.7%. Pre-tax profit grew 14% to A$78.8 million. The company highlighted that its AI solutions business grew more than 100%, and the market rallied on the implication that a small-cap had caught the AI wave.

The numbers are solid. The question is whether the business has fundamentally changed — or whether the margin recovery and cost discipline behind these results are being sold as a transformation story.

The three-line business that makes up A$3.4 billion

Data#3 operates through three segments: Software Solutions (A$2.3 billion), Infrastructure Solutions (A$651 million), and Services (A$412 million). Across all three, it is primarily a channel partner and reseller — sitting between vendors like Microsoft, HP, and Cisco and their enterprise and government customers.

That reseller model creates a gap that matters for understanding what you're actually buying. Gross sales of A$3.4 billion sound large, but statutory revenue — the amount Data#3 actually retains after vendor rebates, pass-through costs, and accounting treatment — was just A$907.3 million. Roughly 73% of every dollar in "gross sales" flows back to the vendors. The company acknowledges that gross profit, not gross sales, is the best view of commercial performance. By that measure, the business retained about A$285 million in gross profit for the year, a gross margin of roughly 8.4% on gross sales.

This isn't unusual for channel partners. It does mean that a 12.7% gross sales headline masks a much smaller underlying economics game.

What actually drove the profit growth

Pre-tax profit grew 14% to A$78.8 million. The driver wasn't a revenue explosion — it was operating leverage. Staff costs rose just over 2% while gross sales grew 12.7%. The internal cost ratio (staff costs as a percentage of gross profit) improved from 79.7% to 77.5%.

In other words, Data#3 managed to grow sales and profit while spending almost the same on payroll. That came from shifting billable headcount offshore and reducing services staff even as it hired more sales specialists. It's the kind of cost discipline that works until it doesn't — management flagged that staff costs will rise up to 10% in FY27, which would add roughly A$2 million in drag on profitability.

The margin story is also more nuanced than the headline suggests. In the first half of FY26, Microsoft changed its channel incentive program, compressing the Software Solutions gross margin from 4.0% to 3.5%. That squeeze alone knocked roughly A$5–7 million of gross profit out of the business, enough to drop the stock 14% when H1 results came out in February.

The H2 recovery — when Microsoft incentives stabilized and infrastructure demand picked up — brought margins back. But what looks like "growth" in the full-year picture is partly a mean reversion. Infrastructure gross profit jumped 18.6%, partly offsetting the Microsoft hit. Software gross profit grew only 7.6% despite 14.1% sales growth, and Services gross profit actually declined 2.2%.

The AI claim, tested

Management said AI solutions grew more than 100%, driven by Microsoft Copilot, Azure AI services, and AI-related consulting. Azure public cloud consumption increased 29%. Security solutions grew 21%. The company also promoted internal AI wins: A$5 million in cost avoidance, 260+ AI agents deployed, and 650+ security events per second processed using AI.

None of this is bad. It's also not a business transformation.

AI solutions are a subset of products Data#3 already resells — Copilot licenses through its Microsoft channel, Azure compute through its cloud business, cybersecurity through its security practice. A >100% growth rate from a small base is encouraging, but management didn't disclose what the base was. If AI solutions were a single-digit percentage of A$907 million in statutory revenue, doubling that figure is meaningful internally but immaterial to the broader earnings picture.

The investment plan reinforces the scale question. Data#3 plans to spend approximately A$2 million in FY27 to build a dedicated AI practice and a sovereign security operations center. That investment will reduce near-term profitability by the same A$2 million. For context, the company holds A$326 million in cash, earns A$10.1 million in interest income, and pays out 90.3% of earnings as dividends. A$2 million is a rounding error in the balance sheet but a clear signal about how seriously this "AI push" is being capitalized.

Compare this to companies whose AI exposure runs through the product layer — selling GPU clusters, building inference infrastructure, or developing AI-native software. Data#3's exposure is at the reseller and services layer. That's a real and growing market, but it's a downstream beneficiary position, not a core AI economics play.

Valuation: pricing a margin recovery as growth

At A$10.72, Data#3 trades at roughly 29 times forward earnings on a basic EPS of 35.16 cents. The stock has gained about 14% over the past 12 months and is approaching its 52-week high. Analyst price targets cluster around A$8.96, suggesting the post-earnings surge has moved the stock past where professional coverage sees fair value.

A 29x multiple is rich for a business with these characteristics: an 8.4% gross margin on gross sales, a 90% dividend payout ratio that limits retained capital, and earnings heavily dependent on vendor rebate structures it doesn't control. The company is debt-free with a strong cash position, which is genuine ballast, but the ROE of 60% partly reflects that low equity base from high payout — not necessarily exceptional operating returns on invested capital.

The 5-year track record is real: 11.6% gross sales CAGR, 16.3% EPS CAGR, and 16% dividend CAGR. That's the kind of consistent compounding that justifies a premium multiple in a stable environment. But the premium also means there's less room for disappointment — the first sign that Microsoft changes incentives again, that services margins don't recover, or that the broader Australian IT market softens further, and the multiple compresses.

What to watch

The investment case here runs through three hinges:

Vendor economics. Microsoft represents a material concentration risk. The incentive change in H1 already proved that vendor policy decisions can swing margins meaningfully. Diversification to non-Microsoft vendors is genuine but the math hasn't fully normalized.

Services recovery. The Services segment — A$412 million in gross sales, down 2.2% in gross profit — is the weak link. Project Services declined 15.3%, People Solutions fell 8.7%. Management expects improvement in FY27, but the drag from economic uncertainty, customer procurement delays, and competitive pricing in maintenance contracts (new contracts averaging 19% margin versus a 27% historical average) will take time to reverse.

AI as category growth, not inflection. The AI narrative is real — enterprises are buying Copilot, Azure AI, and security solutions through channels like Data#3. But the economics are reseller economics. The margin on a Copilot license or an Azure instance doesn't change because the customer is deploying AI. What matters is whether volume growth in these categories is large enough to offset margin compression elsewhere.

The bottom line

Data#3 delivered a solid year of operating leverage and margin recovery. The stock rallied partly on those fundamentals and partly on the AI narrative. The fundamentals are worth paying attention to; the AI narrative is worth pricing at a discount.

For an investor deciding whether to buy in, the question isn't whether the results were good — they were — but whether a 29x multiple on a reseller business that just recovered from a vendor-driven margin shock is the right price. The company has earned the right to a premium through consistency, but consistency alone doesn't justify a multiple that prices in sustained acceleration. The margin recovery has already happened. What the stock needs next is margin expansion beyond where it was before the Microsoft squeeze, and that requires either higher-value services mix or sustained volume growth in thin-margin software resale. Neither is guaranteed.