Visitor Surveys: Spa's DIY Guide

For many years SPA has provided a service to retailers by analysing the Postcodes and spend of customers. Now we offer a similar service to shopping centres, tourist attractions, museums and libraries. And, with cost-control in mind, it's a DIY service. You collect the data. SPA analyses it for you if required.

This guide explains how to do a DIY visitor survey. - download this document as a pdf file for printing (click here to obtain free Acrobat Reader).

Links to each of the examples listed in this section.

So What Do You Want to Know? How about:
	What types of people visit you - and what types don't? We can give you a visitor profile.
	Do patterns vary by time of day, day of week etc?
	How far away do visitors live? We'll give you a breakdown, not just an average figure.
	What is your catchment area? Do some local areas provide few visitors? A map tells all.
	What proportion of visitors come from outside your local area?
	Add your own questions. A simple survey can ask the key questions - about children, services used, or whatever. You don't need a full Usage & Attitude survey for actionable results.

SPA Visitor Analysis reports and maps provide private sector shopping and visitor attractions with hard information on which to base a detailed marketing plan - who to target, where to target, and when to target. Public sector clients more often focus on geographical and social variations in usage of services, the catchments of local facilities (eg branch libraries), and any areas that are poorly served.

Sample Sizes
You would be wasting your money if you did an analysis based on very small numbers with samples too small to be statistically reliable. For this reason, we normally recommend a minimum of 1200 respondents. Allowing for those giving incomplete, incorrect or foreign postcodes, this should yield about 1000 valid postcodes for analysis - the minimum number that SPA would normally profile or split into simple groups. If you're not concerned about statistical significance, you may decide that smaller numbers are worthwhile. When in doubt, go for the largest sample practicable.

1 Design the Survey Sheet - Let SPA do it for you (or DIY )
Survey design isn't easy, so we prefer to do it for you - but it doesn't have to be expensive. For the lowest cost, your survey sheet may be just one line per respondent (see example A1 & A2). We'd design it in Excel using a landscape layout - it's amazing what you can fit in! Collect visitors Postcodes, and we can tell you what kind of people they are, how far they came, and where they live - we can even map them for you. The rest of the line can then be tick boxes, or slots for simple answers (eg amount spent). Use a separate "box" for each letter of the Postcode, and one for each tick. The date of the visit can be in a header, so respondents don't have to enter it.

If you need more than one line per respondent, you can still use Excel to get several entries on a page, but Word may be more flexible. When asking for opinions, it is better to ensure that respondents cannot see each other's answers. This means at least one page per respondent - and higher costs.

2 Register the Survey Sheet with SPA
If you have designed your own survey sheet, you will need to register it with SPA to confirm that we're happy with it. At the same time, we will give you a firm cost for keying in completed survey sheets and analysing it all in due course (if required). We don't want you to use a Survey Sheet with a design fault which makes it easy to misinterpret - or difficult to key-in.

3 Print the Survey Sheet - DIY
When SPA designs your Survey Sheet, we provide you with a paper copy which you can have printed or photocopied, as appropriate (see example A1 & A2). Printing/photocopying is then your responsibility.

4 Ask Visitors to Complete a Survey Sheet - DIY
Once you have duplicated the Survey Sheet, just ask your visitors to fill it in! With a "one line per respondent" survey, have 40-50 sheets comb-bound (or held in a clip-board) on a counter or other writing surface. This allows you to enter header information (dates etc) and to discourage joke entries of various kinds. For longer surveys, you may need a tray or box for completed questionnaires.

Run the survey over a full and continuous calendar period (eg every day for a week, or month). Make sure that inclusion/exclusion of school or bank holidays does not cause distortions. Aim to collect genuinely representative data. If it is impractical for every visitor to fill it in, systematically ask every third visitor, every tenth .. or whatever will work. To compare visitors month on month, you obviously need larger samples than simply to analyse visitors as a whole. Discuss your objectives with SPA to agree an appropriate survey period, sampling strategy and sample size.

5 Key in the Responses - Let SPA do it for you (DIY if you prefer)
To key in survey results yourself, use Excel or Access. Each respondent should be a row, each question a separate column. If your Survey Sheet includes headers (for dates, branch code etc), attach these to all relevant respondents; this is easy using copy and paste. Most problems stem from wrongly keyed Postcodes (letter "O" as zero, "S" as five, etc) or from respondents ticking two mutually exclusive answers. Take care not to enter numeric data in text format or vice versa: it wastes time and increases analysis costs. You can avoid all these pitfalls by asking SPA to key your results for you.

6 Analyse the Data - SPA's speciality (DIY if it's yours too)
If you want simply to cross-tabulate the answers to your various survey questions, you can do this yourself in Access. For smaller surveys, you could even use Excel (the Consolidate Command is invaluable). But if you want to use Postcodes to work out how far visitors travel, what kinds of people they are, and how they compare with the population at large, you may need help. SPA has specialist software for checking Postcodes, attaching grid references, lifestyle codes and Census data.

7 Prepare Tables - SPA or DIY
Producing tables which are attractive and intelligible is not always easy. Unless you're a wizz with table layouts in Word, using Excel may reduce your risk of needing for coronary aftercare services. In Excel you can vary the number and width of columns at will, and reduce printed output to fit the page, so it is generally easier to get what you want without a temper tantrum (see example B1 & C1).

8 Prepare Maps - Let SPA do it for you (DIY if you prefer)
Low cost mapping systems like Microsofts's MapPoint allow you to produce A3 or A4 colour maps of most survey results which include Postcodes. Maps are accurate and attractive (see examples D1, D2 & D3) - sometimes even with street names if you zoom in enough.

Spa can load all your data and map your survey respondents for you (For licencing reasons you must buy a copy of MapPoint from SPA. If you want us to supply maps from MapPoint). Alternatively we can supply you with the survey data files fror you to pull off your own detailed maps when you need them.

9 Prepare Presentation - DIY or SPA
To present your survey results professionally - especially to non-specialists - it is hard to beat the visual impact of a slide show. SPA recommends Microsoft's Powerpoint as an easy way to get over what your survey results actually mean. You can produce the show yourself, or SPA will do it for you.

10 Write Report - DIY (SPA if you prefer)
Writing up a survey isn't difficult, but it takes time and can be expensive to "buy-in". Hence, you may prefer to prepare your own report. As an alternative, SPA offers a report-writing and interpretative service with costs depending on the audience addressed and on the nature and content of the survey itself. Reports can be supplied as Word documents for in-house editing/amendment if required.

Examples & Documentation For a good idea of what results might look like, print out the examples here free of charge. For more detail, we offer a Visitor Survey Pack with somewhat fuller examples and a CD containing:
	Visitor Survey presentation - with advice on keeping down costs down and samples of survey output (some the same as here)
	MOSAIC multi-media documentation software (requires 10mb on your c: drive)
	Outline of other SPA products/services giving a broader idea of analysis possibilities.

For the Survey Pack, send a cheque for £30 (payable to Spa Marketing Systems Ltd) to Visitor Survey Pack, SPA Marketing Systems Ltd, 1 Warwick Street, Leamington Spa, CV32 5LW. If you need a pro forma invoice to raise the cheque, email surveys@spamarketing.co.uk.

Food for Thought
The kind of results you can get from your Visitor Survey depends on the questions you ask.

Visitors' Home Postcodes If you collect visitors' Postcodes, your results could include:
	MOSAIC Profile of visitors:mix of visitors by the 12 MOSAIC Lifestyle Groups and 52 Types
	Distance Profile of visitors: numbers of visitors by distance band
	Visitor Map: a dot map of visitors' home postcodes - useful in defining the catchment area.
	Census Data Profile: age, social class and household profiles for the Census areas from which your visitors come. This is a useful alternative to MOSAIC if you believe that there are certain kinds of areas from which you get very few (or very many) visitors.

Time/Date of Visit If you collect the time/date of the visit, your results could include:
	Time of Day Profile: number of visitors by time of day
	Profiles by Day of Week: MOSAIC and distance profiles split by day of week or weekday v weekend - useful if you suspect that weekday and weekend visitors are significantly different.
Questions Splitting Visitors into Types If your survey include questionswhich split visitors into Types, (eg with/without children, using Service A/B/C/Two+), your results might include:
	Numbers of Visitors by Type: a vital indication of the relative sizes of your main client groups
	Profiles by Type: MOSAIC, Census and distance profiles split by Type of visitor
Questions which Collect Numerical Values If your survey collects numerical data (amount spent, satisfaction 0-5, number in group, etc), your results could include:
	Average value by Type: are some types of visitors bigger spenders/more satisfied?
	Average value by MOSAIC: which MOSAIC types spend least or grumble most?
	Average value by Day of Week: do weekend visitors spend more and come in bigger groups?

Examples of some of these forms of output are provided here.

BUT REMEMBER
Always focus on what you want to achieve. Don't add a question or ask for output just because it will look good. Results which are not relevant to your objectives are a waste of money. If you are interested ONLY in the levels of satisfaction of current visitors, and not where they come from or what kinds of people they are, don't let SPA's enthusiasm for Postcodes lead you to generate visitor dot maps or MOSAIC profiles which - however impressive they may be - are irrelevant to the issues at hand. SPA often uses the power of the Postcode, but there is no hidden agenda to insinuate Postcodes into places where they are not needed. We aim to help you find out what you need to know - at a cost you can afford. We won't achieve that goal if we let you spend money on fripperies!

Your guiding question should always be "Will this tell me something that I need to know?"

Where to Start Just email (fax or telephone) us with the following information:
	a brief summary of the purpose and objectives of your survey
	a list of the kinds of questions you want to ask
	the kind of information you want back
	your draft Survey Sheet (if you're designing it yourself)
	the approximate period/dates when the survey would ideally take place
	your full name, organisation name, address and telephone number

We will then give you a broad indication of likely costs based on our recommendations on sample sizes, layout etc. (If the cost is too high, we'll advise you on how to make savings.) If the suggested cost level is acceptable, we ask you to pay for survey design and registration ... and away you go.

In preparing the above information, you may find it helpful to complete SPA's Survey FAX BACK sheet which covers the key issues. Fax this to us on 01926-450592 if this is more convenient than email.

Visitor Survey Examples
The examples here show what your Visitor Survey might look like. If you're doing the whole survey without SPA's help, copy what you like and alter anything that is inappropriate.

Survey Sheets (examples A1 & A2)
Both examples use one line per respondent, with headers for dates (and branch codes in the multi-site survey). The aim is to collect just a few specific items of information - all that is possible in a one liner. Column headings give instructions and provide extra information for non-locals. Opinion questions ("What did you think of ...?") are avoided: respondents can see each other's answers.

MOSAIC Profile of Visitors (examples B1 & C1)
The report shows the number visitors from each of the twelve MOSAIC Groups. (For large samples, visitors are also split across the 52 MOSAIC Types)

To help interpretation, the mix of visitors by Group is compared with the mix of all GB adults. 24.7% of all visitors are from High Income Families, yet this Group accounts for only 11.8% of all GB adults. The visitor index of 209 indicates that visitors are more twice as likely to come from High Income Family Postcodes as are adults in Great Britain as a whole. By contrast, Council Flats accounts for only 0.8% of visitors compared with 5.7% of GB adults, giving a visitor index of 14. When appropriate, comparison can be with the local area instead of GB. If not required, the bar chart of index scores on the right can be omitted.

Analyses of visitors by distance from home, day of week, time of day, etc are all presented in a similar format - MOSAIC Groups being replaced, by distance bands, days of the week, time slots etc.

Profiles Split by Type (examples B2 & B3)
Any profile can be produced separately for different subsets of visitors - provided the sample is big enough. The example shows MOSAIC visitor profiles split between those who had and those who had not visited shops in the south. Respondents who had not visited South Town shops are seen to be more up-market - largely because south Leamington is less affluent, with higher proportions of postcodes classed as Blue Collar Owners, Low Rise Council, and Victorian Low Status.

Visitor Maps (examples D1, D2 & D3)
A coloured dot shows each visitor's home Postcode. Where two visitors have the same Postcode, only one is visible. Maps from Microsoft's MapPoint have OS quality base map detail (see D1). SPA's version has a simpler base map (see D2), but allows for more flexible choice and colour-coding of visitor symbols. Depending on the data collected, SPA can also present results in the form of colour bar charts (see D3).

What it Costs
SPA's standard service includes the design of your survey sheet and the keying and analysis of results. We provide you with a blank survey sheet for copying and you administer the survey yourself. In most cases, you simply ask arriving (departing) visitors to help with your survey. You can collect as much information as necessary, but excellent results often come from surveys of one line per respondent - focussing on just the postcode and two or three other pieces of information. This helps keep down the cost of the survey and makes it easier to get large numbers of respondents without needing to offer prizes or other incentives which distort the pattern of response.

When your survey is finished, send the completed survey sheets to us for keying and analysis. It isn't possible to quote a fixed price for all surveys. Obviously, if you ask questions which are more difficult to analyse, then your survey will cost more. The Pump Room Visitor Survey includes two open-ended questions probing the reason(s) for the visit. Such questions need to be coded before they can be keyed, and so are more expensive to process than simple Yes/No, multiple choice and cash value questions. As a rough guide, if you collect data from the minimum 1200 respondents, avoid open-ended questions, and your analysis is standard, the minimum cost for the standard service (design, keying and analysis) is about £2200. The figure will be higher if you have a larger sample size and/or more questions, or if you want additional output. You can reduce the cost by designing your own Survey Sheet and by keying in the results yourself. SPA's minimum charge for analysis only is £1500. We can't cover set up overheads for less than that.

The Survey FAX BACK sheet indicates some of the many options available. If you tell us what you want, we can provide an accurate costing for a survey which meets your needs.

back to top>

SPA Marketing Systems Ltd
Leamington Spa, Warks, CV32 6PT
Tel 01926 334978
Email:info@spamarketing.co.uk or Contact us

Please Note: If you wish to PRINT this page out. Click on the page first then select print