WO2008075052A1

WO2008075052A1 - Method and system for performing a survey

Info

Publication number: WO2008075052A1
Application number: PCT/GB2007/004896
Authority: WO
Inventors: Timothy Guy Snaith; Douglas Charles Alan Morris; Neil Jessop
Original assignee: Prime Communications (Uk) Ltd
Priority date: 2006-12-21
Filing date: 2007-12-19
Publication date: 2008-06-26
Also published as: GB0910570D0; GB2457845A; GB0625599D0

Abstract

A method and system for performing a survey of a number of respondents over one or more mobile telephone communications networks is described. The survey is administered by a survey computer on behalf of a surveying party. The method comprises accessing the survey computer by the surveying party and selecting survey data comprising a number of questions as question data and a number of corresponding valid answers as answer data. The question data is automatically provided from the survey computer to each respondent and the corresponding response data from the respondent is received at the survey computer via one or more mobile telephone communications networks, wherein parts of the question data are provided selectively by the survey computer in accordance with the response data received. The results of the survey are communicated from the survey computer to the surveying party.

Description

Method and System for Performing a Survey

Field of the Invention The present invention relates to a method and system for performing a survey over a mobile telephone communications network and in particular for managing the end-end process for creating, purchasing, distributing and reporting on real-time Short Message Service (SMS) and Wireless Application Protocol (WAP) surveys.

Background to the Invention

A survey is a term given to a range of questions presented to a respondent, the responses to which provide insight regarding opinions relevant to those questions. As technology has evolved, surveys have been delivered using a variety of media or channels e.g. face-face, paper, telephone, interactive voice recognition (IVR), web and e-mail. Overall the trend is towards electronic survey transmission at a lower cost and with greater speed.

Known electronic methods predominantly require the respondent to possess the relevant receiver (personal computer with web access, landline telephone, and so on) and to be in a particular location. Additionally, they present the complete survey at one time, meaning that any interruption to the survey leads to loss of response data. Further, these surveys are primarily delivered on behalf of the user by professional services organisations at significant cost.

There is therefore a desire to address these issues and provide a survey at low cost, under the control of the surveying party and which is convenient for respondent use.

Summary of the Invention

A method for performing a survey of a number of respondents over one or more mobile telephone communications networks, the survey being administered by a survey computer on behalf of a surveying party, the method comprising:- a) accessing the survey computer by the surveying party and selecting survey data comprising a number of questions as question data, a number of corresponding valid answers as answer data; b) automatically providing the question data from the survey computer to each respondent and receiving corresponding response data from the respondent at the survey computer, via the one or more mobile telephone communications networks, wherein parts of the question data are provided selectively by the survey computer in accordance with the response data received; and c) communicating the results of the survey from the survey computer to the surveying party.

We have devised an alternative and more effective means of conducting survey processes by enabling a user to manage the end-to-end process without requiring the input of expensive professional service organisations. Where the survey is capable of transmission to the recipient (respondent) for example by means of a mobile telephone and respondent may respond to one question at a time, preferably in a place and at a time of their choosing.

The invention allows the system user (surveying party) to self-manage the survey process for example online using web services, and enables the respondent to respond to those survey questions remotely.

Prior art survey methods do not provide for this holistic self-management of surveys, neither do they deliver requisite survey functionality over beneficial means such as the Short Message Service channel, which to this date have proven to be an intellectual and technical barrier.

The surveying party, also referred to herein as the user or administrator, may operate the survey computer by being positioned at the computer itself or indeed may operate this remotely, for example via the Internet.

The survey itself may be selected by the surveying party either as a previously constructed survey by the surveying party themselves (or a third party), or by actually constructing (generating) the survey data. Therefore a number of surveys in the form of survey data may be stored and accessible upon the remote portal. During the operation of the survey itself, unlike many surveys known in the prior art, in the present case the survey is preferably performed serially in that the questions are provided to each respondent one at a time or in groups such that only part of the question data is provided in response to instances of response data, that is groups of answers or each answer, provided by the respondent. Typically the question data comprises a question and each question generally requires a response from the respondent before a subsequent question is then provided. Some of the question data may of course not include questions and may provide general instructions to the respondent.

The survey data may also comprise routing data, this describing which questions (parts of the question data) to issue to the respondent as a function of respondent answers to one or more previous questions. A comparison is performed between the respondent data and the answer data to determine the subsequent questions to issue. Thus in this sense also, the parts of the question data may be selected in accordance with the response data.

The question and response data are provided between a mobile telephone base station in communication with the survey computer and a mobile telephone operated by the respondent. Preferably the Short Message Service system is used, this providing the questions and answers in the form of text messages. The Wireless Application Protocol (WAP) may also be used in this way. The question and response data sent via the WAP protocol may contain a predetermined identifying code unique to each respondent and survey which is used to identify the respondent to which the response data relates.

It is particularly advantageous to provide the respondent with the ability to choose the protocol (SMS or WAP) which they wish to use. Thus prior to completing the survey, the respondent preferably selects a communication protocol for completing the survey, the said protocol being selected from SMS and WAP. Typically the protocol selection is made in response to an initial SMS message provided to the respondent. This initial message may comprise a WAP link.

When SMS messages are used to contact the respondents, each message is typically issued using a short message peer to peer (SMPP) server. Preferably a number of such servers may be available for use and therefore the method may therefore further comprise selecting a particular SMPP server from a number of such servers based upon one or more predetermined criteria. Such criteria may include server speed, cost, availability, coverage of respondents and so on.

In some cases the survey respondents may be known to the surveying party prior to performing the survey. However, it is particularly advantageous if people can become respondents by initiating contact with the survey computer themselves. Typically in this case the survey is initiated by the respondent sending a trigger message to the survey computer. The trigger message may be an SMS message or WAP message. In some cases when a set of respondents send trigger messages to the survey computer, the computer may select a subset of respondents with whom the survey is performed, rather than each respondent.

Preferably the communicating of the results in step (c) is performed when a predetermined number of respondents have completed the survey. This may comprise a fraction of the respondents and indeed all of the respondents. Updates may also be provided to the surveying party.

In some circumstances, as a result of a comparison in step (b) between the response data and answer data it will be determined that the response data is not in accordance with the expected answer data. In this situation the surveying party may be notified and the surveying party may then amend the response data to bring it into conformity with the answer data.

The surveying party is typically given a choice of when the survey is implemented and this may be effected in terms of a desired start and finish time. Such times may be selected at a particularly advantageous time for the target respondents and the type of survey in question. In general the surveying party will provide the identity of each respondent in the form of identity data. In preferred cases, the identity data are provided by the surveying party by uploading a predetermined template of identity data. However the respondents identified in the identity data may also be selected by a third party or other selection process. When the surveying method is implemented using a telephone network, typically the telephone number of the respondent is provided as part of the identity data.

The identity data may also comprise further descriptive data including one or more of contact data, employment data, geographical data and other socio-demographic data. This may be then used to provide a powerful analysis of the response data thereby providing the surveying party with a detailed understanding of the significance of the responses. Various statistical analysis techniques can be performed using the response and identity data in combination. In order to allow the surveying party to review the data surveyed, the surveying party is preferably provided with the response data during and/or after completion of the survey.

Once the survey has been completed a reward message may be sent to the respondent. This may include a message, barcode or other redeemable voucher.

One or more different mobile telephone protocols (including SMS and WAP) may be used in association with the network(s) to communicate the question and/or answer data. For example the question data may be provided over a first media protocol and the response data over a second, different, media protocol. Different parts of the questions and answer data may also be communicated over different protocols in advanced implementations.

The method is preferably performed by the use of a survey engine which is in communication with the survey computer. The survey engine may actually perform this survey itself once the survey has been selected and initiated by the surveying party. Typically the survey engine is in communication with a message gateway which provides for the selection of particular message gateways from a plurality of possible gateways. A particular gateway may be selected for each message or for a group of messages. Any one survey may therefore use multiple gateways.

Advantageously the surveying party may perform the entire survey remotely using remote access of the survey computer via a website. This may include the use of a mobile telephone via a mobile telephone communications network.

In accordance with a second aspect of the present invention we provide a surveying system comprising a survey computer accessible by a surveying party; and one or more mobile telephone communication networks for providing communication between the survey computer and a number of respondents, the system being adapted in use to perform the method according to the first aspect of the invention.

The system may therefore further comprise a remote device (including a mobile telephone) used by the surveying party to access the survey computer. The survey computer typically comprises a number of sub-systems, one or more of which include a survey management system, a survey store, an account management system and a reporting system. The survey management system provides the interface for the user. The survey store provides a database of possible surveys and their associated data such as question data, answer data and identity data. A background survey processing system may also be provided to communicate with the communication network. This may be a separate system from that of the survey computer and this may contain a survey engine and a message gateway processor for selecting message gateways from a plurality of possible choices. The gateways may comprise SMPP or HTTP servers. The system may also comprise an inbound server adapted to receive responses from the respondents. The survey engine is preferably adapted in use to generate an individual software process for each respondent, the process being arranged to match inbound and outbound messages between the survey engine and the respondent.

Where WAP functionality is used the system may also comprise a Wireless Application Protocol (WAP) server to send and receive questions and response data when the survey is completed using the WAP protocol.

A scheduling server may be provided to select and issue surveys to the respondents according to a predetermined time schedule. A database may be used to store the response data. Preferably a database controller is provided for controlling the rate of transmittal of the response data into the database.

Each of the individual parts of the above system may be distributed over a network.

It will be appreciated that the system is largely computer-implemented and many of the method steps may be performed by suitable computer program code means. These may be stored on a computer-readable medium.

BRIEF DESCRIPTION OF THE DRAWINGS

An example of a survey method and system according to the present invention is now described with reference to the accompanying drawings, in which:-

FIG. 1 is a block diagram of a system for conducting a real-time SMS survey according to an example of the invention;

FIG. 2 is a flow diagram explaining the operation of associated website software; FIGS. 2A-2E provide various screen shots associated with Fig. 2;

FIG. 2F shows a draft recipient upload template;

FIGS. 2G-2L shows further screen shots associated with FIG. 2;

FIG. 3 is a flow diagram illustrating the operation of the survey engine;

FIG. 4 is a flow diagram of the report viewer system; FIGS. 4A-4C shows screen shots associated with FIG. 4;

FIG. 5 is a diagram of the accounting system; FIG. 6 is a flow diagram of the message gateway system; FIG. 7 is an overview of the system apparatus; and, FIG. 8 is a further flow diagram of the method.

Detailed Description of the Preferred Embodiments

The following eight figures and descriptions illustrate an example of 1) the system overview, 2) the survey creation mechanism, 3) the survey distribution mechanism, 4) the message gateway mechanism, 5) the report viewer mechanism, 6) the system administration mechanism, 7) the distributed system apparatus and, 8) the survey process.

To illustrate the innovation more clearly, the following description is an example of a typical survey process according to the invention using a survey management process wizard. This is performed typically over the Internet where a user (surveying party) sitting at a remote computer accesses an Online Survey Management System (OSMS) computer (survey computer) which operates software for performing the method. Further details of the apparatus and method for implementing the survey are then described.

With reference to Figure 1 , the OSMS is shown schematically (left hand side). This comprises a survey store in the form of a database, a survey management system S100 in communication with the survey store, an account management system S400, in communication with the survey store and a reporting system, again in communication with the survey store. The OSMS computer is further in communication with a background survey processing system which oversees the execution of the survey. This comprises a survey engine S200 and a message gateway processor S500. The message gateway processor S500 is under the control of the survey engine S200 and both sends and receives signals to mobile telephone networks so as to oversee the delivery and receipt of short message service (SMS) messages, commonly referred to as "text messages". The respondents of the survey in this example are the individual holders of the mobile telephones to which the text messages are delivered and from which the respondent messages are received. Further details of the survey engine S200, Message Gateway Processor S500 and other related hardware are discussed in more detail later in association with Figure 7. A parallel system using a choice of Wireless Application Protocol (WAP) and SMS is also discussed with reference to Figure 7 and Figure 8 later.

A user who is desirous of performing a survey of mobile phone customers initially logs on to the OSMS computer and enrolls through a registration process described later in Figure 5. This is performed under the control of the account management system S400.

Reference is now made to the flow diagram of Figure 2, in association with which an explanation of the operation of the portal is given.

Having logged on to the OSMS computer portal (embodied as a website), the user is firstly presented with the new survey page as illustrated in Figure 2a. In this example this is the user's first time on the website and therefore the software determines that the survey is a new one at step S100a in Figure 2. Should the user have already visited the site and wanted to reuse a previously written survey they could select a survey from one of their previous surveys that are automatically stored (S104c and S100b).

As shown in Figure 2a at step S101 the user then completes the survey name, their company name (or that of the company they are conducting the survey on behalf of), provides an easily memorable survey reference number and a brief description to aid recall for future identification of results. As this is a new survey and is not an expert survey (S102), the user then, by clicking "next", is taken to a "creating and editing questions" (S103b), an example page being shown in Figure 2b.

Had the user chosen to use an 'expert^* survey they would have be taken to a similar questions page (S103a), but would have been unable to edit/amend the questions. Note that an expert survey is one in which pre-defined questions and a predefined survey structure are used.

Since this is not an expert survey, the user clicks on a "new question" option, which adds a question to the main viewing screen and provides a question number e.g. "Q1". Here there are three tabs: o "General", where the user adds the question wording and the question type (Opt out, Multi-choice, rating, free text (Figure 2b); o "Answers", where the answer options are created for that question (shown in

Figure 2C); and, o "Routing", which enables intelligent routing of specific respondent options to specific future questions (see Figure 2d)

Having created their survey, the user clicks on test questions, which launches a web browser page within which they can check the logic of their questions, answers and routing without having to launch the survey and discover errors in a live environment (Figure 2e).

When the user is satisfied that the survey will work as intended and has therefore explored the functioning of the survey in terms of how the survey proceeds as a result of different responses to the questions, the user is now ready to click "next' and progress to add the list of intended recipients for that survey.

As this is their first time on the site in this example, the user downloads a recipient template (S105). This is a template setting out information regarding the recipients of the survey, namely the intended mobile telephone respondents. The template is completed by entering the mobile telephone numbers and all other relevant reporting details into the spreadsheet shown in Figure 2f. Once completed this is then uploaded to the system (S106b). The next time the user visits the site this upload template will be automatically available to them in their stored recipient lists as shown in Figure 2g. If the user has previously visited the site and uploaded one or more lists of recipients then they may select one of these lists at S106a.

The user has now created their survey, and provided a list of recipients. By clicking "next" (S107), the cost of the survey is calculated based on these two variables and presented to the user. The cost is then compared with the credit within the user's account (S108). If there are sufficient credits in their account (S 109) the user is then allowed to continue (S110a) and is provided with options regarding when the survey is to be performed and their "survey notification" requirements (S111). The notification requirements allow the user to be updated regarding the progress of the survey. In this case, both email and SMS are chosen as means for such communication (Figure 2h).

By selecting all of the potential notifications, the user is alerted initially when the survey goes live and then when they receive 25%, 50% and 75% response rates and when the survey ends. Each of these is designed to keep the user up to date with results.

Once the survey is scheduled, the user is routed to the home page, from where they can access all of the functionality of the portal (see Figure 2i). Note that the user can save the survey they are working on at steps S104, S107, S110a, for later recall. This is automatic and doesn't require user intervention.

The survey has a progress report in a "live surveys" area accessible from the main page of the website where its details can be viewed (Figure 2j). Once it goes live (typically at a predetermined time), it is sent for immediate distribution, to the survey engine (S112).

Once the survey is live, additional completion data is available to view, including any recipient anomaly notifications (Figure 2k). These recipient anomaly notifications indicate an error in a recipient's response, this being where the recipient has given an unexpected response such as one which is not one of the possible answers defined earlier. The user is notified and they may then go to the "fix anomalies" page to see if they can remedy the response and re-trigger that individual respondent's survey (see Figure 2I). This process is summarised in steps S208, 208a, 208b, 208c of Figure 3 to be described below.

The survey continues to run, with each response from each individual respondent triggering the next appropriate response from either the system, the user (where anomalies are fixed), and the respondent. An overview of the technical operation of the survey engine throughout this process is shown in Figure 3. Further details are discussed later in association with Figures 7 and 8.

The operation of the survey process is controlled by the survey engine. The survey begins at the appointed time (S200a) with the user being notified if they have requested this (S201 ). The first message is sent to the first respondent at step S203 via the message gateway. Unless the message does not require a response (S204a), the recipient survey is placed on hold at S205. Unless there are no more response messages expected from the respondent (S206a) the process moves on at S206 to check for the response of the respondent received in a response message queue (used for all recipient messages) via the message gateway. The response is compared with the valid responses dictated by the user (S208) and the response, if valid, is added to the survey results (S209). Thereafter the survey proceeds (S210) provided the survey time has not expired (S211 ). The process then returns to S202 where, following a check that there are indeed more messages to be sent to the first respondent (S202) the second message for the first respondent is then sent via the gateway at S203 and the process repeats. This process is performed for all respondents in the survey effectively simultaneously (although the messages are actually queued in practice).

As will be appreciated, if an invalid response is given at step S208 by a particular respondent then this is flagged to the user administrator at S208b and corrected by the user at s208c. The corrected response is inputted into the survey results at S209. Once the survey run time (as determined by the user) has expired then this is determined at S211 and the survey ends with a completion message being sent to the user at S212. During the survey messages are sent to and received from each recipient such that at any one time multiple recipients may be taking part in the survey. Of course some respondents will answer each of the questions rapidly and for these the survey will be completed relatively quickly. For others who are unable to complete the entire survey at once (due to other commitments, moving between poor areas of mobile reception and so on, or because an erroneous answer is given), the completion of the survey may take appreciably longer.

The user can view their survey results in real-time, both during and after the conclusion of the scheduled survey period. This is shown in Figure 4.

The user first selects the survey for which they wish to review the results (see Figure 4a) at step S300a. The survey results are presented one at a time (S301 ,S302), with the user clicking "next question" to proceed (see figure 4b below). Because the user uploaded additional respondent data in the upload template (See figure 2f) the user can now filter the results against these variables by clicking on the 'Analyse by^* option (S303) shown in Figure 4c. This enables powerful reporting whilst utilising short succinct questions/surveys. The user created their own survey and therefore they will not be able to utilise the expert survey benchmark facility (S304). The bench mark facility allows comparisons between the data received from the respondents and that received from other sets of respondents.

As will be appreciated, the present invention therefore provides many advantages over known surveying methods and provides the surveying party with unparalleled control over their survey at low cost without the need for an intervening survey-operating company. They can however elect to download all of the results (S305a) in a .csv file for use by other software. The download contains all of the questions and all of the scheduling information in order to retain detailed survey records with the data sheet. . An overview of the accounting system is shown in Figure 5. The top part of Figure 5 shows the hierarchical structure according to which the system is organized. The lowest level of the hierarchy is the Basic Administrators which are the individual users of the system. These are the surveying parties who are the end users interested in the results of the surveys. They typically pay a fee to use the system to perform surveys. They have the ability to manage their own accounts with the system. An intermediate level in the hierarchy is occupied by Reseller Administrators. These have the ability to perform surveys themselves as well as control the pricing and use of the system by the Basic Administrators. The highest level is occupied by the Super Administrator which has complete control over the system and represents for example a company which provides and manages the entire system as a commercial service. The Super Administrator may therefore allocate credits to any user, generate global expert surveys for the benefit of the individual user customers and also manage the message gateways used to provide the surveys (as discussed below). The lower part of Figure 5 summarizes the administration rights.

The message gateway system S500 is shown in more detail in Figure 6. This enables the administrator to select a preferred gateway S500a through which to distribute the survey. This facility enables the administrator to select gateways according to their resilience, quality of distribution, cost per message, and any stated preferences of the user. Gateways can be thought of most simply as different mobile telephone providers within the present example. More than one gateway can be used for each survey and the protocol rules for each selected gateway are selected at S501. After performing a credit check to ensure that no unrecoverable costs will be incurred at step S502 a message is sent at step S504 following a check that this is to be performed S503. Incoming messages are also processed by the gateway at step S503a.

Figure 7 shows in more detail a preferred arrangement of the apparatus for implementing the survey system. This particular design allows each of the components of Figure 7 to be distributed at different locations and networked. The preferred implementation of Figure 7 consists of the survey engine S200 which stores the surveys submitted by the administrator. A shotgun server S602 (scheduling server) in communication with the survey engine provides surveys to the survey engine S200 and instructs the survey engine S200 when to send the surveys according to the schedule required by the survey administrator. As mentioned earlier, the survey engine S200 provides the ability for surveys to be conducted by WAP (Wireless Application Protocol) or SMS (Short Message Service). For SMS surveys, the survey engine creates an individual client thread S601 (a software process) for each respondent which controls each of the inbound and outbound messages to the remote respondent.

The survey engine ensures that the outgoing question data and incoming response data are correctly matched up to the respective client thread using the caller line identity system which contains the mobile telephone number of the respondent. The client thread communicates with the messaging gateway S500 to send and receive the survey questions and responses. An inbound server S605 receives the inbound messages from the respondent and transfers them to the survey engine S200. A journal manager S603 (database manager) and a journal queue S604 are arranged in communication with the client threads. They receive the response data from the client threads and transfer them to the database S607 at an appropriate time. In the case of a survey using the WAP protocol, a WAP server S606 communicates directly with the survey engine S200 and the journal manager S603. The WAP server is operative to send and receive the questions and responses sent using WAP.

Figure 8 shows an example method of performing a survey using the apparatus of Figure 7. At step S700 the survey starts. At an appropriate time, the shotgun server

S602 sends a survey to the survey engine S200 according to the schedule created by the administrator. At step S701 the survey engine S200 generates an individual client thread S601 for each respondent. The client thread contains respondent identifying data in the form of caller line identity data (the mobile telephone number of the respondent) for each individual respondent so that inbound and outbound messages can be matched with the correct respondent.

The first message to send to the respondent is generated at step S702. This is a special initial message that allows the user to have the choice of either completing the survey using SMS or WAP. The initial message is always sent using SMS. The message is sent to the messaging gateway S500 at step S703 for delivery to the respondent. To achieve this the client thread S601 connects to the messaging gateway S500, and requests and selects a gateway for the client thread to use. The messaging gateway S500 typically uses this specific gateway for issuing subsequent messages until the survey is finished or a fault is detected. The messaging gateway S500 uses mobile telephone network specific gateways and the respective network recognises the data as an SMS message.

To send the SMS survey data the survey engine S200 sends the SMS data via the messaging gateway to a particular SMPP server (short message peer to peer server) chosen from a number of such servers. The SMPP server sends the SMS message to the mobile telephone through the Internet and ultimately through the mobile telephone network. HTTP servers can additionally or alternatively be used to send the SMS messages; however, SMPP servers are preferred as they have an improved performance over HTTP servers.

The survey engine S200 typically uses the same SMPP server for each respondent and in each case waits for acknowledgement from the server before continuing. The coverage of the SMPP servers is limited geographically. However, it is possible to connect to multiple SMPP servers so that a large proportion of the population can be surveyed. The messaging gateway S500 has a routing capability to select the SMPP server from the recipients telephone number, for example according to country code. Multiple SMPP servers are used by the system; ensuring that if one SMPP server goes offline then the SMS messages will still be sent and received. Preferably, the fastest SMPP server is used although if this is not possible then the next fastest SMPP server is chosen. Other criteria in addition to speed may also be used to select the servers, as discussed in association with Figure 6. Importantly, when respondents reply to the SMS survey, the reply may be received by any of the SMPP servers.

When the client thread receives an acknowledgment from the outbound server the thread goes into a receive mode. At step S704 the respondent makes their choice of protocol by selecting WAP or SMS and replies to the message accordingly. If the respondent chooses to complete the survey using SMS then at step S705 the SMS response and the caller line identity is received at the inbound server S605. The inbound server S605 monitors multiple inbound messaging gateways (SMPP or HTTP servers). When an inbound message is received, at step S706 the caller line identity of the received message is used to assign the inbound message to the correct client thread S601. If the response received is not a valid response to the question sent then the anomaly procedure is initiated (S208a). The client thread for the particular respondent then communicates with the journal manager S603 and at step S707 the data received from the respondent is transferred to the journal manager. The journal manager "drops" all of the data (both valid and invalid) received from the respondent into the database S607. The journal manager greatly speeds up the overall operation of the system since it is solely responsible for the queuing of data into the database and this function is therefore taken away from the client threads. The journal queue S604 allows the data received from the messaging gateway S500 to be held in a queue until the database is ready to receive it.

At step S708 having processed and stored the last response of the respondent, the client thread then proceeds with the next question in the survey by sending this to the to the respondent using the SMS protocol according to the method described earlier. Steps S705 to S708 are repeated until each of the questions have been sent and a corresponding response received. It should be noted that, once a question has been sent via SMS a new question will not be sent until a response to the original question has been received. If no response is received within a predetermined time limit a reminder message can be sent or the question re-sent. At step S709 the survey ends after all of the questions have been sent, responses have been received and the response data transferred to the database.

Returning now to step S704 at which the respondent makes the choice between WAP and SMS, if at step S704 the respondent chooses to complete the survey using the WAP option then the second route shown in Figure 8 (right hand side) is used. At step S710 the respondent receives a web link to allow the respondent to access the survey which is embedded in the initial SMS sent to the respondent at S704. The respondent uses this link in step S711 to send a response to the WAP server S606. The WAP server S606 replaces the inbound server S605 in the survey process. The WAP server and the SMPP servers that send the SMS messages run independently of each other. The WAP server is only initiated in this case due to an inbound request from the respondent. When a response is detected by the WAP server a message is sent to the survey engine S200 to "kill" the client thread in step S712 and the survey engine assumes the survey is proceeding via the WAP option. Therefore, no more SMS messages are issued and the survey engine plays no further part in the survey process. In the case of WAP, the caller line identity information is replaced by a unique code which is generated so as to be mobile phone and survey specific. The code is attached to the inbound and outbound WAP messages so that when the responses are received by the WAP server they can be assigned to the correct survey and respondent. The WAP server is similar to a Web server and the respondent converses with the WAP server as they would with Web server.

Once the initial WAP message has been received from the respondent, in step S713 the WAP server S606 communicates directly with the journal manager S604 and transfers the data received to the database S607. When a response has been received by the WAP server the next link is sent if required and the responses transferred to the journal manager and database. This process of steps S714 and S715 may be repeated a number of times using different question data (depending upon the complexity of the survey). At step S709 the survey ends after each of the questions have been issued and the respondent answers received and transferred to the database.

In either case, using SMS or WAP, once the response data has been received from a number of respondents it may then be accessed by the user of the system.

As illustrated in Figure 7 a survey administration system communicates with the database S607 and provides functionality including an anomaly fixer, survey results viewer and survey scheduler.

A number of optional modifications to the above system are contemplated. For example, it is possible for either the respondent or the survey administrator to choose whether to send the survey using the SMS or WAP protocol depending on survey content and the location of the respondent. However, the initial contact with the respondent is preferably via an SMS message whether the WAP or SMS option will be chosen to complete the survey.

A highly advantageous option with this invention is that a survey can be generated without the initial outgoing message from the survey engine S200 or a pre-constructed client thread S601 for the respondent. For example, a survey can be requested by sending a key "trigger" word to the inbound server S605. The server recognises this as a trigger and transfers the message to the shotgun server S602 which then checks this trigger keyword with the keyword of the surveys stored on the database. If the keyword matches one of the surveys on the database a corresponding survey is sent to the mobile phone and a new client thread is generated for the user and the survey continues as previously described. If the message sent to the shotgun server S602 does not correlate with a keyword of one of the surveys then the message is determined to be spam and discarded. This function of the invention is useful for generating new respondents to the survey without the need for outbound messages. In particular the user interested in performing the surveying of respondents is not required to have the contact details (telephone number) of the respondent in question prior to the survey beginning.

Surveys can also be triggered by new respondents using the WAP function in a similar manner to the SMS option. For example, if a user goes to the web link on their mobile phone without having received the initial text message a keyword can be inputted and the survey sent back either as an SMS as described previously or with another web link provided for WAP completion of the survey. It is also possible if required for one respondent to trigger the survey on another person's mobile telephone. If the keyword and the mobile telephone number of the respondent are sent to the inbound server S605 then, in this case, it will trigger the survey from the shotgun server S602 and send it to the target telephone and not the initial sender's mobile phone. This then allows the details of the new respondent to be entered into the database S607 and a new client thread generated as previously described. In either case, triggering is also valuable in that data identifying the respondent can be stored, allowing such a respondent to be contacted for future surveys.

A further option is to personalise the survey to the respondent in question. In this case part or all of the data received in previous questions can be inserted into subsequent questions. For example, if the first question is "what is your first name?" then the second question can address the respondent by their name. This adds to the attractiveness of the survey from a respondent perspective.

Another option to personalise the survey to the respondent is to attach personal data ("strings") of the respondents to the client thread so that personalised data can be incorporated into the questions sent when using both the WAP and SMS survey methods. An example of incorporating the "string" would be to address the respondent by name immediately rather than asking their name in the first question. This data is uploaded to the system (S106b) on creation of an upload template as described earlier (Figure 2f).

When using either communication method (WAP/SMS) to complete the survey, there are two types of selectivity for providing the question data. The first is a survey where there is no flexibility in the questions asked. The questions generated by the survey administrator are sent serially on receipt of a response from the respondent. The second is to generate the questions depending on the previous answer to the questions. For example, if the first question is "do you read newspapers?" and the answer is no then the second question can be directed to "do you read magazines?" rather than "do you read broadsheets or tabloids?" Routing of questions is also possible on receipt of answers to questions that require a rating as their response. For example, a different subsequent question can be sent depending on ratings that are equal to or less than, equal to or greater than, equal to, less than or greater than the ratings in the question. Such a rating is typically numerical and routing of the data may depend upon the answer value or the particular numerical range within which the answer falls.

In some cases reward messages can be sent on the successful completion of a survey. For example, a multimedia message with a bar code that can be scanned at a point of sale terminal could be sent or a text message could be sent with a live web link to a reward or a printable voucher on a website.

If desired by the administrator the respondent's data can be kept anonymous from the administrators so that the names and the telephone numbers of the respondents are not matched with the data received. A flag can be set in the survey so that the mobile telephone identity data is not recorded and the survey answers are only seen in the reporting of the results. On initiation of the survey a question can be sent to the respondent to ask them if they would like their responses to remain anonymous. This will then exclude the identity data of those who select that option whilst retaining the identity data for those who do not select the option.

It is also possible, if required, to restrict the number of respondents to a trigger survey such that only a subset of respondents is allowed to perform the survey. An example of this would be if the administrator only wants to record the opinions of the first 1000 respondents and then anyone else who sends the keyword to the survey engine will receive a thank you message in response. To encourage respondent interest in a survey the average results of the survey can also be sent to the respondents when the survey period is completed.

It is also possible to implement a completely mobile version of the administrator website, enabling the generating, sending and monitoring of the surveys to be fully mobile. Whilst the example discussed above exemplifies the use of a mobile telephone network to implement a survey using the short messaging service protocol, those skilled in the art will appreciate that this can be extended readily to the provision of the surveys across multiple communications networks using one or more media protocols.

Claims

1. A method for performing a survey of a number of respondents over one or more mobile telephone communications networks, the survey being administered by a survey computer on behalf of a surveying party, the method comprising:- a) accessing the survey computer by the surveying party and selecting survey data comprising a number of questions as question data, a number of corresponding valid answers as answer data; b) automatically providing the question data from the survey computer to each respondent and receiving corresponding response data from the respondent at the survey computer, via the one or more mobile telephone communications networks, wherein parts of the question data are provided selectively by the survey computer in accordance with the response data received; and c) communicating the results of the survey from the survey computer to the surveying party.

2. A method according to claim 1 , wherein the survey is selected in step (a) by selecting a predetermined survey in the form of survey data stored previously.

3. A method according to claim 1 or claim 2, wherein the selection of the survey data in step (a) is performed by the surveying party generating the survey data using survey generation software.

4. A method according to any of the preceding claims, wherein in step (b) one or more questions are provided to each respondent serially such that only part of the question data is provided in response to each instance of response data provided by the respondent.

5. A method according to claim 5, wherein the part of the question data comprises a question and wherein each question requires a response from the respective respondent before a subsequent question is provided.

6. A method according to any of the preceding claims, wherein the survey data further comprises routing data and wherein the parts of the question data are selected in accordance with the routing data and the outcome of a comparison between the response data and the answer data.

7. A method according to any of the preceding claims, wherein the question and response data are provided between a mobile telephone base station in communication with the survey computer, and a mobile telephone, using a short messaging service text message.

8. A method according to any of the preceding claims, wherein the question and response data are provided between a mobile telephone base station in communication with the survey computer, and a mobile telephone, using the wireless application protocol (WAP).

9. A method according to claim 8, wherein the question and response data sent via the WAP protocol contains a predetermined identifying code unique to each respondent and survey, the method further comprising using the identifying code to identify the respondent to which the response data relates.

10. A method according to any of claims claim 7 to 9, wherein prior to completing the survey, the respondent selects a communication protocol for completing the survey, the said protocol being selected from SMS and WAP.

11. A method according to claim 10, wherein the said protocol selection is made in response to an initial SMS message provided to the respondent.

12. A method according to claim 11 , wherein the initial message comprises a WAP link.

13. A method according to any of the preceding claims, wherein when SMS messages are used, each message is issued using a short message peer to peer (SMPP) server.

14. A method according to claim 13, further comprising selecting a particular SMPP server from a number of such servers based upon one or more predetermined criteria.

15. A method according to any of the preceding claims, wherein the survey is initiated by the respondent sending a trigger message to the survey computer.

16. A method according to claim 15, wherein the trigger message is an SMS message or WAP message.

17. A method according to claim 15 or claim 16, wherein when a set of respondents send trigger messages to the survey computer, the computer selects a subset of respondents with whom the survey is performed.

18. A method according to any of the preceding claims, wherein step (c) is performed when a predetermined number of respondents have completed the survey.

19. A method according to any of the preceding claims, wherein when the result of the comparison in step (b) reveals that the response data is not in accordance with the answer data, the surveying party is notified.

20. A method according to claim 19, wherein, following notification, the surveying party amends the response data to bring it into conformity with the answer data.

21. A method according to any of the preceding claims, wherein the surveying party provides the identity of the respondents in the form of identity data.

22. A method according to claim 21 , wherein the identity data includes a telephone number to which the question data are sent.

23. A method according to claim 21 or claim 22, wherein the identity data are uploaded to the survey computer in accordance with a predetermined template.

24. A method according to any of claims 21 to 23, wherein the identity data comprises additional descriptive information describing the respondents.

25. A method according to claim 24, further comprising analyzing the response data received from the respondents in accordance with the descriptive information.

26. A method according to any of the preceding claims, further comprising, on completion of a survey, sending reward message to the respondent.

27. A method according to any of the preceding claims, wherein the surveying party is provided with the response data during and/or after completion of the survey.

28. A method according to any of the preceding claims, wherein the survey is controlled by a survey engine, in communication with the survey computer.

29. A method according to any of the preceding claims, wherein the surveying party interacts with the survey computer via a website.

30. A method according to claim 29, wherein the surveying party interacts with the survey computer using a mobile telephone via a mobile telephone communications network.

31. A surveying system comprising:- a survey computer accessible by a surveying party; and, one or more mobile telephone communication networks for providing communication between the survey computer and a number of respondents, the system being adapted in use to perform the method according to any of the preceding claims.

32. A system according to claim 31 , further comprising a remote device for use by the surveying party to access the survey computer so as to perform a survey.

33. A system according to claim 32, wherein the device is a mobile telephone.

34. A system according to any of claims 31 to 33, wherein the survey computer comprises one or more of a survey management system, a survey store, an account management system and a reporting system.

35. A system according to any of claims 31 to 34, further comprising a background survey processing system to communicate with the communications network.

36. A system according to claim 35, wherein the background survey processing system comprises a survey engine and a message gateway processor.

37. A system according to claim 36, wherein the message gateway processor is adapted to select between a plurality of message gateways.

38. A system according to claim 37, wherein the message gateways comprise short message peer to peer (SMPP) servers.

39. A system according any of claims 36 to 38, further comprising an inbound server adapted to receive responses from the respondents.

40. A system according to any of claims 36 to 39, further comprising a Wireless Application Protocol (WAP) server to send and receive questions and response data when the survey is completed using the WAP protocol.

41. A system according to any of claims 31 to 40, further comprising a scheduling server to select and issue surveys to the respondents according to a predetermined time schedule.

42. A system according to any of claims 31 to 41 , further comprising a database for storing the response data.

43. A system according to claim 42, further comprising a database controller for controlling the rate of transmittal of the response data into the database.

44. A system according to claim 36, wherein the survey engine is adapted in use to generate an individual software process for each respondent, the process being arranged to match inbound and outbound messages between the survey engine and the respondent.