A so-called independent lab report into competing failover systems has shone a spotlight onto the murky practice of companies paying for tests and then using the results to promote their product.

The results of a test on three products - WANSync, Double-Take for Exchange and Neverfail - found that Neverfail had performed best in general. "Neverfail provided the best protection for many of the potential causes of failure (hardware, network, software, performance, configuration, disaster), something competitive offerings from XOsoft and Double-Take did not provide," said the test's principal researcher, Tom Henderson from ExtremeLabs.

This report was then extensively promoted by Neverfail, both by its press relations company, which put out a press release and supplied copies of the report to journalists, and at a press conference by Neverfail's CEO Neil Robertson. In every case, Neverfail stressed that the report was independent. The report itself makes no mention of who paid for it or of any potential conflicts of interest.

When Robertson was then specifically asked at a press conference in London on Tuesday if Neverfail had paid for the report, he replied: "Absolutely not. There's a period where you're sweating to know what they're going to say but it was completely out of our hands."

The truth however is that Neverfail did pay for the report. In fact, it paid for the entire lab test exclusively and in full.

Questions subsequently put to the company by Techworld elicited the fact that Neverfail had first approached ExtremeLabs and Tom Henderson. But the company still refused to state whether it had paid for the tests. "Even though the test was brought to him by Neverfail we really didn't have much to do with it," explained Adam Schaeffer, a spokesman for Neverfail in the US.

Then one of the companies that was featured in the report as inferior to Neverfail's product responded, accusing Neverfail of being less than truthful and questioning the report's validity. "This was a report commissioned by our competitor and does not appear to us to be an accurate representation of our product's capabilities," said Ian Masters, UK sales director at Double-Take Software. "There are several findings which are simply not true that should bring into question the validity of the entire report." Masters confirmed that his company did not contribute in any way to the report.

Eventually, when quizzed for a third time, Neverfail admitted it had paid for both the tests and the report, and sought to play down its CEO's comments. PR director for Neverfail, Ann Taylor, told Techworld: "We paid for the test. Neil knew the testing was being done. He just didn't quite understand that we had paid for it."

The company apologised again the next day: "Neil apologies for his comments," we were told. "When he made those remarks he wasn't aware that Neverfail had paid for the tests. He had no intention of misleading anyone."

Meanwhile, the man who had carried out the tests was frank about Neverfail's involvement and about the system of sponsored lab reports that has grown up in the industry. Asked if Neverfail commissioned and paid for the test, Tom Henderson of ExtremeLabs told us: "That's correct."

He wouldn't reveal the actual cost, however, saying he was obligated not to. What about the avowed independence of his position? "In my business I probably do 15 of these kinds of reports a year; some for vendors, some for competitive analysis. ... Our vendor's preference is that we don't pull punches. If it hadn't turned out well, you never would have seen the report."

You can read the report, available on the Neverfail website, here.