Banner Advertising

Portal Sites

Banner advertising is best used on sites where high traffic is expected. These sites are commonly called portals because they represent the first stop for the web surfer. The most popular form of portal is the search engine. Essentially there are two ways to place banner advertising on a search engine, a) run of site (ROS) and b) keyword associated. Run of site means your banners will be displayed randomly across the site. Keyword associated means that your banner is displayed when certain key words are searched, given the specialist nature of the service/product this is likely to be more effective.

Both of these banners are most commonly charged at Cost per thousand page views or CPM, (not per thousand clicks through). Therefore it is especially important to ensure that the audience is appropriate for the specialist product/service we offer.

Words of warning: Because the searcher is more interested in the results of their search than your banner, the banner must be very attention grabbing. I would recommend therefore that (depending on in-house skills) an external agency be engaged to produce a series of banners for the different keywords purchased. I would also recommend that banners are only purchased when you have established your listings on search engines. The creative cost here will obviously vary according to the agency but could be between £2,000 and £6,000. This is justified because of the critical importance of effective banners in your plans for success. A banner is typically animated to help gain attention, making it more complicated than a standard graphic to produce.

By sponsoring keywords you can identify that the viewer is searching for a product/service you are offering. If they seek that service/product there is a strong possibility they are looking to purchase.

The cost of this approach varies according to search engine (these figures include some $ conversions at a $1.6 exchange rate): -

If your audience is very targeted, this can have benefits from an online marketing point of view. Primarily it means that you can seek out more specialised web sites which don't get the traffic of the main sites but which will provide you with the audience you seek. In this case however the viewer has not selected themselves as seeking your service by searching for keywords, you have selected them for readership or audience compatibility.

Magazine titles often form the focus for the online community in that field. Because of the lower viewing figures you may be able to advertise at much reduced rates while still hitting your target audience and in locations which will directly raise your profile with existing contacts.

The Newsweek site could be an example although it isn't specialist enough for our purposes. Here the rates vary from £37.50 per CPM when you buy 100,000 views down to £25 per CPM for 1 million.

Another approach is to conduct an Internet search for key words you associate with your service/products. Sites rating high should be considered for advertising or if appropriate sponsorship.

The numbers (working out the coverage you need)

Let us say that your target in order to double your online revenue is £1/2 million and you know that currently 40% of online business is generated by new customers and 60% by existing customers. I believe that an online campaign will benefit both of these groups in terms of volume. However because it is targeted at newcomers we will see an increase in the proportion of new customers in this mix. Perhaps this will make the equation 50/50. Therefore the target for new business revenue would be £250,000. If we assume we are only talking about a product valued at £1,500 we will need to generate 166 online sales to achieve your objective.

If you have never conducted online advertising or banner advertising before you need to make some assumptions.

The equation needed to calculate the CPM you must reach is: -

x=(y.a)b

Where: -

x = to the number of new customers needed

y = to the size of CPM to reach

a = The percentage of clicks through the banner and

b = The percentage purchase of those that do click through

If we assume that 5% of the people who see our banner click through on it. (Based on the fact they have self-select, by searching for your keywords and the ease with which the viewer can 'click through'). Of these 5%, 1% go on to make a purchase within three months of their first visit (if not on their first visit - see explanation below). The equation would be: -

166=(yx0.05)0.01

where: -

166 = the number of new customers we need

y = the unknown size of audience we must reach

0.05 = the number of 'y' audience clicking through to our site

and

0.01 = the 1% of the result who will purchase within three months.

Obviously both the 5% and 1% estimates have to be tested, they are definitely your weak points in this calculation. However, the numbers are conservative so they should form a solid basis for conjecture.

The result of the calculation gives us a reach of 332,000 to achieve our objective.

n.b. These figures are exclusively based on banner advertising generated new viewers and assumes only one product is purchased per customer. By enhancing our activity in other areas, particularly search engine listings we would expect to see additional viewers being channelled to the site from these streams. The benefit of this will be to fortify these projections and allow some flexibility in results. It is also worth remembering that if we need to generate twice the income we would double the coverage to 664,000 impressions.

As with the example of Yahoo! above, because of the specialist nature of your market it could take sometime to reach your 332k CPM figure. By spreading your spend across all main search engines and by purchasing several keywords you will improve the time in which this will be achieved.

The ROI (return on investment) on these calculations is excellent and could represent anything from 7:1 to over 30:1. These figures are based upon a single purchase per customer and do not reflect any 'lifetime value' calculation. We might for example surmise that 60% of new customers would repeat purchase and this benefit could also be set again the cost of recruitment.