The inefficiency of the Web is legendary.
The story you are about to read is true. The names have been changed or suppressed to protect the incompetent. Editorial comments are in italics. The rest actually happened.
10/25/2000: Client filled out project brief. (Hooray, the project is off and running!)
11/15/2000: Agency presented media concepts and strategy documents. (Moving fast!)
12/01/2000: Client put plans on hold while reviewing budgets.
2/26/2001: Agency presented final plan (version 5) and buying guidelines now that client has an approved budget.
3/7/2001: Sent RFPs to 57 sites - small budget in specialized market results in a smaller list than normal.
3/15/2001: 38 of the 57 sites have responded to the RFP by deadline, 5 have asked for extensions, 15 - no response.
advertisement
advertisement
3/15/2001: Sent follow-up email to the 15 sites that did not respond to RFP, 2 emails returned as undeliverable that had been delivered previously.
3/19/2001: Called the 5 reps that did not respond to follow-up email to RFP; 2 phone numbers no longer working.
3/23/2001: Submitted preliminary buy recommendation to client. Client requests we look at five more small sites that they have become aware of. (Client starts thinking about the reality of the campaign.)
3/24/2001: Can't contact 2 of the sites, 1 doesn't sell advertising, the remainder want cash in advance and will make no audience guarantees.
3/25/2001: Client removes 2 sites from the final buy when they realize one of their sales people has entered into preliminary biz-dev talks directly with those sites. (This usually comes with a related budget reduction.)
3/26/2001: Major portal decides to walk from the buy over Terms & Conditions (client very upset), which puts agency at risk. (Note that the Ts&Cs in contention were originally agreed to in V 1.0 of the IAB/AAAA's Ts&Cs established at a meeting which a representative of this portal attended and apparently agreed to.)
3/26/2001: Two sites will not accept 3rd party ad servers, even though the original RFP stated that sites should not submit proposals unless they accept 3rd party ad servers.)
3/26/2001: Buy is officially one week behind schedule!
3/28/2001: Client approves final buy; schedule starts... sort of.
3/30/2001: Tags on 2 of the client's 7 websites are broken, their technicians are swamped and they will try to get them fixed in the ‘next couple of days.’
3/30/2001: Tell client we’ll have to take the two banners that feature those sites out of rotation until tags are fixed.
4/2/2001: Out of 23 placements only 7 are actually up and running on time.
4/4/2001: Email drops are scheduled for tomorrow but we will have to postpone because the client didn't get the copy ready and has to run it past their legal department before release. (This often happens when client insists on saving a buck by doing creative "in-house." But guess who gets blamed when things do not go according to plan?)
4/5/2001: Noticed that the "Click here" image in the HTML email is Java-based and many email services cannot support Java, so it has to be removed from the source code in the ad running on one of the email buys in the campaign. (See note about client doing creative "in-house" above.)
4/6/2001: Major search vehicle text links not up yet due to "site technical issues."
4/8/2001: Search engine text links finally up.
4/15/2001: Mid-campaign report reflects 6 situations where delivery is seriously under estimates. Contact sites and they all say that things seems to be picking up and they’re sure they will deliver by the end of the schedule.
4/16/2001: Cost per acquisition nowhere near the goal on one third of the sites. Recommend to the client that we cancel and move the money to the better performing sites. Also contact all sites that are marginal to see if they can help improve upon the results. Get good cooperation here.
4/17/2001: Under-performing sites all hold to one-week cancellation but do agree to provide targeted bonuses to improve performance.
4/18/2001: Only one of the best performing sites has additional appropriate inventory.
4/18/2001: Three sites that have performed well feel that they will need another week to deliver, client agrees to the extension.
4/19/2001: Found out that a targeted placement on a major search engine was lost when site accidentally deactivated the client's tags.
4/20/2001: Mid-campaign adjustments all made.
4/20/2001: Search engine placement back up.
4/27/2001: Nothing went wrong this week!
4/30/2001: Trying to pull sample campaign reports; unable to get the post-impression data. Contacted 3rd party ad server.
5/1/2001: Waiting for all sites to complete schedules. Unbelievably, half of the placements have been completed although results for only a few are as good as desired despite mid-campaign tweaks.
5/4/2001: Still no post-impression tracking data available.
5/10/2001: Campaign is over with 5 sites unable to deliver the entire inventory purchased. Told them we will have to pro-rate based on actual delivery.
5/18/2001: Provide preliminary final report although post-impression data still unavailable.
6/6/2001: Due to 3rd party ad server database problems client still doesn’t have the final report and the campaign has been over for almost a month.
6/12/2001: Provide final report and analysis to the client. Results were mixed. The agency felt that learning from this campaign and optimization on future campaigns would reflect even greater success -- results showed that the client could acquire customers for less than half their traditional cost per acquisition. (Why is it that the Internet is held to a higher standard?)
6/15/2001: Client advises agency that they are not sure they will continue with any Interactive campaigns. Despite some obvious success, new CFO has stated that he "just doesn't believe in this interactive stuff." (Wonder what he’s going to do about the annual biz dev deals and how those are performing?)
7/10/2001: Accounting reports that they have not received invoices from 5 sites and have discrepancies to resolve on 75% of the invoices received to date.
10/25/2001: Last billing problem finally resolved, sent refund to client for sites that did not fully deliver. Client asks why the buying fees are not being returned on the media that did not run. Explained to the client that our contract stipulates that fees are based on the amount of the original purchase. (After all, we did the work and more.) Explained that we have found no way to ensure that sites always deliver and that just as we don't expect incremental payment from the client if we select sites that over-deliver, we can't be penalized when there are under-deliveries.
The notes above came from actual emails and contact reports on the business. The client was a small division of a major company in a category where Interactive media has proven to be a major success factor.
In the end, we learned lessons we’ve learned before:
1) You can never charge enough to handle a small test buy
2) NEVER take on a campaign that runs for less than three months, giving you a chance to truly optimize and get great, not just good results.
3) Internet budgets of a number of major companies remain at the mercy of top management people who don't believe.
Notes to self:
* "Hire" clients who believe
* Take on bigger, more long-term efforts
* Raise fee structure on short, smaller campaigns
* Don't negotiate with sites who have not signed Ts&Cs - push harder to get them signed with RFP submission
* Put SLA (Service Level Agreement) wording into next 3rd party server agreement
relative to termination for non-performance.
* Continue to work hard to standardize this exciting but difficult industry.
Karen T. McFee is Executive Vice President and David L. Smith is President of Mediasmith, Inc., the Integrated Solutions Media Agency based in San Francisco and New York.