up  over  a  number  of  pages,  displaying  perhaps  ten  results  on  every  page.  The  problem 

with  this  approach  is  that  very  few  users  are  willing  to  look  at  more  than  two  or  three 

pages`  worth  of  results.  If,  as  in  the  case  of  the  Golden  Pages  site,  the  businesses  are 

sorted alphabetically by business name by default, Acme Plumbing Co will appear on the 

first  page  by  virtue  of  nothing  else  other  than  having  a  name  starting  with   A`,  whilst 

McSweeney  &  Son  Plumbers  and  Worthington  Plumbers  and  Fitters  are  buried  several 

pages further down the list, dramatically reducing their visibility to any potential clients.  

Dedicated  search  engines  such  as  Google  have  gone  to  great  trouble  developing 

sophisticated algorithms to sort the potentially hundreds of thousands of results generated 

by a search for, say,  Britney Spears , so that the user rarely has to look beyond the first 

page  of  results.  A  Golden  Pages  style  service,  drawing  from  a  much  smaller  and  more 

tightly  defined  database,  has  less  room  for  manoeuvre.  Results  will  be  pulled  straight 

from a relational database table and there is no real practical or fair way to generate and 

store the kind of dynamic metadata to feed something like the PageRank algorithm.  

The idea of user ratings has potential, but  peer review  on the Web does bring with it a 

number  of  major  issues.  So  as  to  prevent  either  unfair  promotion  or  sabotage  of  a 

particular business, either human editors would have to verify every rating as it came in, 

or  users  would  have  to  create  some  sort  of  account  on  the  service,  allowing  spurious 

ratings to be easily traced back. The first option is potentially expensive and the second is 

unlikely to be taken up by any more than a handful of users; few people would go to the 

trouble of submitting their personal details to an online service when all they want is to 

find a telephone number. The general question of reliable peer review was deemed to be 

outside the scope of this project. It is nonetheless a fascinating and very pertinent issue, 

and one worthy of further research. 

Faced  with  this  problem  of  not  being  able  to  meaningfully  sort  results,  the  only  fair 

option  (from  the  point  of  view  of  Mr.  Worthington,  the  plumber)  is  to  have  all  of  the 

results on one page. From a user`s perspective, it means less clicking and, assuming the 

user  were  to  otherwise  load  every  single  page`s  worth  of  ten  results, less loading time. 

The argument that the user would not bother to look through every one of the sets of ten 

serves  only  to  highlight  Mr.  Worthington`s  disadvantage.  In  this  author`s  admittedly 

subjective  view,  scrolling  down  one  long  page  is  a  good  deal  less  unappealing  than 

wading through a series of individual pages which  

20