Visitors come to your site and don't buy on their first visit.It often takes 6-7 visits before they buy. So, how do you gettheir business if they don't come back? For that matter, how doyou get them t
Visitors come to your site and don't buy on their first visit.
It often takes 6-7 visits before they buy. So, how do you get
their business if they don't come back? For that matter, how do
you get them to come back?
One answer is to run a contest. Not just any old contest will
do! It must be crafted to keep them returning again and again.
Lots of smart people will tell you that running a contest will
help but they don't tell you how to do it. Here's a blueprint
you can alter to suit your purposes.
Every aspect of the contest was handled by CGI scripts and a
couple of web pages. Even selecting the winner was done by a
script. This final script simply located the persons with the
most referrals. It drew one at random if there was a tie for
first place.
One of my clients came up with the basic idea and we developed
it together as a team. People would get one "entry" into the
contest when they signed up for one or more of his newsletters.
They would get another "entry" for each person they referred.
It was hugely successful. During the 60 days of the contest,
over 45,000 new people had subscribed to my client's
newsletters. They were coming back to his site over and over
again. Orders for ezine ads poured in.
At the end of the contest period a week-long cruise vacation to
Montego Bay, Jamaica was awarded to the winner. Generally, if
the prize is valued at over $500, some authorities in your state
may expect to be notified. There are various requirements; check
with the appropriate state agency for details.
We posted a complete set of rules for the contest on the Web
site. We announced the contest itself as well as the current
front-runners in each issue of my client's newsletters.
Each person entering the contest was assigned a Personal
Identification Number (PIN). The PIN was used to connect them
with their referrals. Next, a person was given the chance to
tell their friends about the contest.
We made everything as easy as possible for the contestants. All
they had to do was supply the email address and name and the
scripts did the rest.
They'd be shown the message that was about to be sent to their
friend. We were making every effort to be honest with them and
not cause any ill will.
When their friend(s) entered the contest, using a PIN we gave
them, their "referral count" would be be increased. To win, a
contestant had to get the most referrals. In the event of a
multi-person tie for first place, the software would
automatically conduct a random drawing.
During the contest, each contestant was able to use a special
Web page to monitor the contest. They'd see the top- ranked
contestants and the number of referrals each one had made. They
could log in with their PIN and see their own referrals, too.
They were able to come back to the site at any time and enter
new referrals. We added a small form that would look up their
PIN in case they'd forgotten it. They could send their own
emails to friends to tell them about the contest. They could
place ezines ads giving contest details and a PIN to use.
Keeping up with all those contestants, who they referred, how
many they referred, etc., was a job for a database engine. Since
we had MySQL on a server we controlled, we used that. The
scripting was done in Perl.
Two Perl modules were needed: DBI.pm and the MySQL module for
Perl. The modules are chunks of Perl code that let a script send
commands in SQL (the generic language of databases) from the
script to the MySQL database engine (program).
You can use a different database; you just need the the
database-specific module so Perl can "talk" to it through the
DBI module. DBI is needed regardless of which database you use.
Specific modules exist for Oracle, PostGres, etc.
We verified the email addresses of all the winner's referrals;
we would have to disqualify a contestant if any of the addresses
were faked - and award the prize to someone else. We designed
the code so that any attempts to cheat actually decreased your
odds of winning - and it was all automatic!
Thousands of names and email addresses are in the database. The
referral counts have been reset to zero and we are ready to
launch another contest at any time we choose.
The programming effort, for the most part is all done. We'll
just have to edit the contest page(s) and the rules page to
reflect the new prize and the new "start" and "end" dates. We
can easily get *at least* as large a subscriber boost with our
next contest!
Here's the bottom line: You'll get more visitors, and they'll
visit again and again. You get to show them your offer enough
times to win their business. No matter what product or service
you're selling, an automated referral contest is something you
should seriously consider.