It seems to be common knowledge that 80 or 90% of small businesses fail in the first year. I've heard this statistic enough times that it seems as if it must be a fact proven multiple times and written in stone somewhere. But the problem is its not really a fact.
I mentioned a while ago that Matt at YFNCG posts in Statistics Shmatistics about the wide variation in statistics cited about failure rates of new small businesses.
In his article he cites several sources with statistics on the small business failure rate that are all over the place.
Matt cites sources with failure rates as high as 95% and as low as 25%:
“Nearly half of small businesses fail within a year…90 to 95 percent of small businesses fail within five years.” - Acording (sic) to this wisegeek.com article, the stats come from the Small Business Association, but they don’t say what reports were used or who in the SBA reported them."
"In an article on Small Business Trends Scott Shane presents stats from his book that show only 25% of businesses fail after the first year and 50% are still going strong after year 5."
I just found another source myself from Dr. Gregory B. Murphy a professor at the college of business at University of Southern Indiana where he explains why the numbers might be so far off:
"Conventional wisdom regarding small business failure rates is based on the assumption that if a small business can be identified as new in a given year and cannot be identified five years later, it’s a failure."
So if the business is in a list one year and not on the list the next year they count it as a failure.
"Just because a small business is no longer listed in a given source doesn’t mean it is no longer there. A number of sources that list small businesses are based on compiled yellow pages. Other sources require the small business owner to pay a fee and/or provide company information to be included in their listings. So what happens if a small business is no longer listed in a source because the owner no longer wants to pay the listing fee or provide the requested information? You guessed it. If that source is being used to track the small business, it probably will be counted as a failure — even if nothing else about the business has changed."
So part of the problem may just be poor quality data.
Other reasons why the 'failure' rate data isn't necessarily good is what they call a failure:
"If an entrepreneur successfully starts and grows a business and then sells it to another company at a nice profit, should that business be counted as a failure because it no longer exists as a separate entity? Clearly not. If a near-retirement professional starts a consulting company to cap off his or her career and does so successfully and then retires, closing the business four years later, should that business be counted as a failure? Ironically, very successful small businesses may be mistakenly labeled as failures."
So if you sell a business then it could end up in the 'failure' bin. Clearly that's not right.
The high numbers in the 80-95% range don't even pass a sanity check. Think about how many businesses you know of in your town. Consider all the restaurants, bars, barbers, furniture shops, car dealers, and all the other small businesses. Then think about how many of them went out of business in the past year. Is it anywhere near 80-90%? I'd bet its almost always not nearly that high.
If you hear someone say that 80 or 90% of businesses fail in the first year then take it with a grain of salt.