Analyzing Market Potential for an Art Gallery
Project Overview
The purpose of this project is to identify potential markets for an art gallery in the Halifax Regional Municipality, Nova Scotia. 409 census Dissemination Areas (DAs) were selected within and around the city. Three variables were chosen from the 2013 Canadian census data (derived from the Esri Canada Business Analyst Standard 2013 data set) for use in suitability modelling. Fifteen competitor businesses were geocoded, and classified into their respective market segments. Suitability Index ranking logic and variable weight rationale is discussed. The impact of weighted variable suitability on primary market distribution and competitor location is compared against unweighted distribution of variables.
// GIS for Business
ArtPerFurn represents money spent on art as a percent of total spending on furnishings. As the business is looking to sell pieces of art, this variable seemed to be very applicable. Values ranged from 3% to 16%, with larger values being ranked higher on the suitability index. The greater average spending on art within a DA, the better for an art gallery.
Variable Selection
The Income variable represents average household income within a DA. It is assumed that households with higher annual incomes will have more disposable income for use on decorating their homes. As such, higher suitability rankings were given to DAs with larger average incomes.
PerTarPop represents the target population as a percent of total population within each DA. Persons aged 40-69 were chosen as the target population, as it is assumed that this age group will be established in their career enough to have more money to spend on art. Any persons younger than 40 may not have enough money, or desire, to purchase art for their homes. Any persons older than 69 may have already sufficiently decorated their homes to not spend significant amounts on more art. DAs with greater percentages of the target population are given higher suitability ranks.
Distribution of Unweighted Markets and Competitors
Suitability Index rankings for each of the three variables were summed to determine each DAs’ summed suitability (SumSuit). Suitability Sums were classified into Primary, Secondary, and Tertiary markets to determine optimal locations for an art gallery. Out of the 409 DAs, 65 were chosen as Primary areas at 206 km2. All variables were used equally to establish these results, to show unweighted market class distribution. Only DAs with a Summed Suitability greater than 12 were chosen to represent primary markets.
Competitor art galleries within Halifax-Dartmouth were found through the Yellow Pages website, and geocoded in ArcMap. Surprisingly, none of the competitor locations were within DAs classified as primary markets. In fact, all but one of the competitors were located in tertiary market areas. Whether this is an issue with the variables (not enough information), an unaccounted socioeconomic factor (tourist flux, cheaper storefront rent), or a variable weighting issue (determined in the next step) is unknown. It could also be possible that the model is still correct, and art gallery owners did not factor this information into their decision making process when locating their business, and instead decided to cluster around other competitors (follow the herd).
Total area (km2) per unweighted market category
Variables were weighed (see table to the right) in order to redistribute the significance that each plays in the model. PerTarPop was given a much lesser weight, as age isn’t the best indicator for who will be purchasing art (everyone can appreciate art). Income and ArtPerFurn were given more importance, as people with more money and a greater desire to spend money on art were deemed to be the best indicators for choosing the location of an art gallery.
By applying these weights, a larger number of DAs were classified as primary markets (66 more). As such, less areas were classified as secondary and tertiary. Although the weighted market distribution changed considerably, the classified competitors did not. Only one extra competitor moved from tertiary to secondary- the remainder were all located within tertiary market areas. Thus, variable weights are not the reason for competitors clustering in tertiary areas. It appears as though more analysis is necessary to determine the reason for competitor clustering in apparently suboptimal locations.
Weighted Variable Distribution Index
Total area (km2) per weighted market category
Distribution of Weighted Markets and Competitors
Created by Sean Thibert
Datum: WGS 1984
Projection: Mercator Auxiliary Sphere
Geocoded Competitor Addresses: Yellow Pages website (YellowPages.ca)
Census Data: Esri Canada Business Analyst Standard 2013 data set
Disclaimer: These maps are for educational purposes only, and should not be used for navigation.
