Coordination in a franchise channel

This study aimed to replicate in South Africa a survey by Guiltinan et Bl (1980) on the effect on coordination in a franchise channel of: Franchise authority; franchise influence; helpfulness of communications; automony; and uncertainty. The results will be of value to both franchisors and franchisees. In addition the results in South Africa are very similar to the United States study. This has important implications for United States companies wanting to expand their operations to South Africa. S. Afr. J. Bus. Mgmt. 1984, 15: 150-154


Introduction
The realization during the late 1950s and 1960s of the importance of the franchisor/franchisee relationship and all its implications in terms of synergy between the efforts of the franchisor and the benefits to the franchise system as a whole, on the one hand, and the efforts of the franchisee and the benefits to himself, on the other, led to a considerable amount of attention being focused on this relationship.The relationship is also important for other reasons, as pointed out by Vaughn (1982): (a) To motivate franchisees to perform at a high level in terms of both marketing and internal operations; (b) to facilitate the adoption of new ideas; (c) to obtain new ideas from franchisees, and ( d) to prevent or minimize legal problems relating to antitrust, trademarks and trade names.Starting early in the 1970s, a nwnber of studies were published specifically in the area of distribution channels, most of the studies viewing them as behavioural systems.

Research Areas
Despite the implications of Rosenberg & Stem's (1970) paper that minimiz.ation of channel conflict required closer coordination within the channel, empirical channel research did not focus on the issue of coordination, but rather on the questions of power, conflict, and satisfaction.
Later Robert F. Lusch (1976) published a study which gave limited support to the assumption (frequently found in the marketing and retailing literature) that conflict in a channel of distribution affects the operating performance of the members of the channel and/ or channel efficiency (Mallen, 1967).His study used the very interdependent relationship I» tween manufacturer and dealer in the franchise channel for the distribution of automobiles.Conflict was measured as dealers' perceived frequency of disagreement with the manufacturer over twenty previously validated issues.Performance was measured by return on assets and asset turnover.For two of the five single-franchise channels measured, the regressionanalysis results supported the hypothesis that increasing levels of channel conflict were associated with lower dealer operating performance.
For the other three, the lack of significance may have been due to potential errors in the measurements of conflict and operating performance obscuring any relationship inherent in the data.In any event, causality was not established in the case of the two significant franchise channels, and it may have been S. Afr.J. Bus. Mgmt. 1984, 15(3) that dealers with low operating performance tended to blame it on the franchisor, hence giving cause for frequent conflicts.
The first study published that dealt with coordination and the factors affecting it -only in 1980 -was that of Guiltinan Rejab & Rogers (1980).

'
This study aimed to replicate in South Africa that of Guiltinan et al (1980) on the effect on coordination, in a franchise channe~ of: Franchisor authority; franchisee influence: helpfulness of communications; autonomy; and uncertainty'.

Scope and Methodology
An anonymous questionnaire was mailed, with a covering letter and self-addressed envelope, to all the twenty-seven individuals who are franchisees of Juicy Lucy.(Although the chain has over sixty outlets, a number of them are company-owned, and some of the franchisees own more than one outlet.) Following an interval of four weeks, phone calls were made to non-responders to follow up the reasons for non-response.The majority gave reasons such as 'Too much trouble' 'Couldn't be bothered', and a few were away on leave.Ther; were no responses among both responders and non-responders that indicated they had any problem understanding or completing the questionnaire.This confirms the similar findings of Guiltinan et al's (1980) pretest.
In all, there were thirteen responders and fourteen nonresponders representing a response rate of 48,2%.This compares very favourably with the 12,30Jo response rate that Guiltinan et al., achieved from their sample of franchisees, which was over forty times the size of the sampling universe in this study.Nevertheless, it must be remembered that fastfood chains in the United States of America serve a target market of over 230 million people, whereas that of Juicy Lucy in South Africa is virtually only 28 million, hence the enormous difference in the sizes of chains.

Hypotheses
These were identical to those framed by Guiltinan et al ( 1980), which was dictated by the fact that this was a replication study.The hypotheses were developed on the basis of the literature available to them at the time.The authors' own survey of the literature as well as publications subsequent to this study (for example Vaughn, 1982) failed to reveal any indication that these hypotheses are no longer appropriate and it seems safe to conclude that no changes to the hypotheses are necessary.The hypotheses therefore are (quoted directly): H1: The greater the degree of franchisor authority, the greater the level of coordination perceived by franchisees.
H2: The greater the degree of influence exerted by the franchisee, the greater the level of coordination perceived by franchisees.
H3: The greater the degree of autonomy perceived by the franchisee, the lower the degree of coordination perceived by the franchisee.~: The less the amount and importance of environmental uncertainty, the lower the degree of coordination perceived by the franchisees.Hs: The more the information emanating from a franchisor is perceived as being helpful, the greater the level of coordination perceived by the franchisee.Guiltinan et al. (1980).

Questionnaire
The questionnaire was designed to measure franchisee responses on six areas using a five-point Likert-type (agree/disagree scale): (a) The helpfulness to the franchisee of franchisor communications, both printed and personal.(b) The belief that the franchisor has the authority to specify various functions in the outlets.(c) The franchisee's belief that he could influence franchisor decisions or guidelines on various functions.(d) The franchisee's perceived autonomy as measured by his felt-ability to influence his outlet's sales volume and operating costs.(e) The franchisee's perception of the amount and importance of (environmental) uncertainty regarding sales and profits.
(0 The franchisee's perception of the amount of coordination existing between him and the franchisor on the various functions.The functions used to elicit the responses in the areas of authority, influence, and coordination were: Food preparation; menus; stock control; operating hours; advertising; display and presentation; employee training; accounting and pricing. These differ from Guiltinan et al (1980) only in respect of 'sales promotion' -which is little used in the Juicy Lucy chain and therefore inappropriate, and was replaced with 'display and presentation' -and 'pricing' -which Guiltinan et al (1980) did not consider, but which represented an important issue about which the Juicy Lucy franchisor required information.
In the area of communications (printed and personal), responses were elicited with respect to cost control, competition, as had Guiltinan et al (1980), but due to inapplicability of 'sales promotions' the question of 'new developments' was used instead.
The elicited responses in the areas of autonomy and uncertainty are indicated above in (a) (and (e)) respectively.
Guiltinan et al (1980) established the reliability of the multiitem scales measuring work coordination, information helpfulness, authority, influence on franchisor, autonomy, and uncertainty using the coefficient alpha.(See Peter (1979) for discussion and justification.)Based on sources such as Nunnaly (1967), the internal consistency of the scales appeared to fall easily within the established acceptability criteria.
Guiltinan et al (1980) used principal components factor analysis to assess the independence or interdependence of the first five of the six response areas, hoping to show that they could be considered independent variables in the study, with 'coordination' being the dependent variable.They were only partially successful in this attempt.
In this study the small number of franchisees in the universe studied precluded the use of techniques such as factor analysis, but it was expected that the use of correlations on appropriately chosen groups of data would yield sufficient indications of the franchisor/franchisee relationship in the chain studied.

Limitations of the Study
Certainly the small sample size and the small universe size for this study, and the attendant implications, making statistical significance almost totally beyond reach, must be the most important limitation of this study.This came about for reasons inherent in the nature of the study (being a replication) and because of population and market-size realities, about which little can be done.
Although it was theoretically possible to choose a franchise chain in South Africa with more outlets (Wimpy is the largest, with over 300 outlets), in practice this would have removed much of the motivation for the study.

Results and discussion
In order to analyse the choices on the five-point Likert-type scale, integer scores from O to 4 were assigned to the responses 'strongly disagree', 'disagree', 'not sure', 'agree', and 'strongly agree' respectively.
Because of the small sample size, many of the statistical techniques, which would normally be applied to establish levels of significance or confidence intervals, could not be meaningfully used.In addition, tests of significance of the difference between two means (like the t-test) rest on an assumption of independence between the two groups of data, and there was no justification for an assumption of independence in a situation like this.For example, there was no reason to assume independence between perceived coordination, as measured on the function 'operating hours', and perceived franchisor authority over operating hours.
As a result of these problems, it is almost impossible to make statements about the results with any measurable or estimable degree of confidence.Therefore, findings are stated in terms of apparent trends indicated by the correlation coefficients calculated on the data.In some instances conclusions are drawn -rather speculatively -on the basis of differences that stand out of the tabulated data.For example, in Table 1 -which shows the pair-wise correlation for coordination and authority as measured by each of the nine functions -the only two functions with means below 2,00 are 'stock control' and 'operating hours'.The low scores for coordination as measured by these two may be interpreted as franchisee perception that there is, indeed, little cooperation on stock control and operating hours.This might in fact be true, but if we take account of the standard deviation for the stock control scores being 1.01, and if we assume an underlying normal distribution, then we can say with 66,711/o confidence that the true population mean for that score lies between 0, 76 and 2, 78 (i.e. 1,n, approx 1,01), (or with 9511/o confidence that the true mean lies between O and 3, 79 (i.e.l, 77, approx 2 x 1,01 ).This clearly goes outside the limits of normal statistical statements and thus any conclusions drawn can only be termed possible trends.
In the same way, one might argue that the low means for authority as measured by 'stock control' and 'operating hours' imply little belief on the part of franchisees that the franchisor has the authority to specify stock control and operating hours.But this too can only be stated as a possible trend.

Table 1 Pair-wise correlation for coordination and authority
The large standard deviation for coordination and authority as measured by the function 'accounting' may be taken as an indication that perceptions of franchisees in this area differed widely.Conversely, the lower standard deviations for 'food preparation' and 'display and presentation' (on both authority and coordination) could be indicative of greater uniformity of opinion among franchisees.This is probably so, as these two are areas in which the franchisor, studied here, definitely does specify in considerable detail what will be done and places emphasis on these areas when checking stores.Food preparation is also perceived as the area in which most coordination takes place and over which the franchisor has most authority.This also makes sense as the franchise 'stands or falls' on the quality of its food products ('its basic market offering').
In Guiltinan et al's (1980) factor analysis, authority emerged as two separate measures, one comprising authority on food preparation and menus only, the other comprising the remaining six functions; this split could be explained by the fact that food preparation and menus represent the 'basic market offering', which is strategic and different to the other functions measured.This is in keeping with the comments (above) regarding 'food preparation' and 'display and presentation' in this study.
In order to assess whether authority might be split into two separate measures in our sample, pair-wise correlations were calculated for coordination and authority as measured by all nine functions; by the remaining eight with 'menu' excluded; by the remaining seven with 'menu' and 'food preparation' excluded; and by the remaining seven with 'food preparation' and 'display and presentation' excluded.As can be seen from Table 2, the correlation excluding 'menu' yielded exactly the same coefficient (0,80) as when all nine functions were excluded; the correlation between coordination and authority was at its highest (0,84).This finding is similar to that of Guiltinan et al (1980) and may support the idea that the influence of those functions seen to be, or made out to be, concerned with the 'basic market offering' is of a different nature to the influence of the other more mundane functions.This might well be because the 'basic market offering' related to the strategic reasons for joining the channel in the first place, and the different quality of the 'food preparation' and 'display and presentation' functions may explain why, taken together, they correlate authority and coordination rather weakly.
As the correlations in all but the last line of Table 2 are all very close, it was considered justifiable to use the combined scores for all nine functions for coordination against which to measure the other four areas (apart from authority) being  studied.Table 3 shows these correlations.Franchisee influence correlated reasonably, but not very strongly, with coordination.As can be seen from Table 4, the correlation coefficient of 0,45 is very similar to the 0,459 of Guiltinan et al. (1980).Autonomy was related even less to coordination than in Guiltinan et al's (1980) study -in both instances not significantly.Helpfulness of franchisor infonnation correlated fairly well with coordination in this study, more so than in Guiltinan et al. (1980).There was nothing to choose between printed and personal infonnation as the difference in their two correlation coefficients was negligible.Guiltinan et al (1980) had hypothesized a negative correlation between uncertainty and coordination (Hypothesis 4) but found to their surprise the opposite relationship, and explained that this may be because 'the franchisor has done an effective job of uncertainty absorbtion'.In the present study, the hypothesized negative correlation was found, but only meaningfully in respect of uncertainty in franchisee prediction of 'sales' and 'profits'.(For importance of predicting ability of sales and profits and for all four of these combined, the correlations were not meaningful.)Perhaps this means that the franchisor has not 'absorbed uncertainty', although it does not seem necessary to find an additional justification for a confirmed hypothesis, beyond the theory and background on Considering the hypotheses individually, it may be tentatively concluded, on the basis of the trends indicated by the correlation coefficients, that only some of the hypotheses are upheld.Hypothesis 1 appears to be fairly strongly supported, much more so than in Guiltinan et a/'s (1980) study.Hypothesis 2 is not very strongly supported, as in the Guiltinan et al., ( 1980) finding.Hypothesis 3 was not meaningfully upheld at all, which mirrors Guiltinan et al. (1980).Hypothesis 4 was reasonably strongly supported, although only in respect of uncertainty in franchisee prediction of 'sales' and 'profits'.This was contrary to Guiltinan et al. (1980).Hypothesis 5 was more strongly supported than in the prior study.

Summary of results
The results of the study can therefore be summariz.ed.In the channel studied: (a) Franchisee perception of franchisor authority appears to be quite strongly correlated with perceived level of coordination.(b) Franchisees' perceptions of the extent to which they can influence the franchisor's decisions or guidelines on various issues do not appear to correlate much with their perceived level of coordination.(c) Franchisees' autonomy perceptions appear to be unrelated to their perceptions of coordination in the channel.(d) It appears that franchisee uncertainty of the 'sales' and 'profits' predictions is negatively correlated with their perceptions of coordination.(e) Perceived helpfulness of franchisor infonnation also appears to be positively correlated with franchisee perceptions of coordination.(Guiltinan et a/'s (1980) correlation coefficients are stated to 3 decimal places because their larger sample size justifies it.In the present case it was not justified to use more than 2 decimal places.)

Conclusions and recommendations
As with Guiltinan et al. (1980), the study does add weight to the argument that viewing contractual channels and cooperative and interacting partnerships is likely to increase perfonnance of the individuals making up the channel and of the channel as a whole.This is because such a view encourages channel participants to indulge in the kind of behaviour patterns that are now known to improve coordination and communication and to affect the various factors involved in channel relationships.(And also encourages avoidance of those behaviour patterns which work against channel perfonnance.) The findings of this study indicate a greater effect between franchisor authority and franchisees' perceptions of coordination than that revealed by the Guiltinan et al (1980) study.This might be due to the less regulated legal and operational environment for franchisors in South Africa than that in the United States of America, allowing franchisors to be more authoritarian in their approach to franchisees, although this cannot be proved.
The lack of a very strong correlation between franchisee influence on franchisor decision and/ or guidelines on various issues and perceived coordination should not encourage franchisors to take decisions concerning the franchise system in isolation.The indication is certainly there of a positive correlation and recent American experience with Franchise Ad-1S4 visory Councils appears to support the idea that making franchisees feel part of the decision-making process has positive benefits.These can be expected to include not only better performance, but also new developments on the part of the franchisees.
Although the fmding of both studies did not fmd any significant relationship between perceived coordination and franchisee autonomy, there is evidence from other sources (e.g.Vaughn, 1982) that personality differences, earnings and previous employment are worthy of attention when it comes to selecting franchisees.Vaughn (1982) recommends that franchisees be recruited among people earning less than $10 000 per year and those who have previously been working in sales.These groups appear to be 'more often satisfied with their franchises than others'.
This study managed to confirm, at least in part, the hyoth~ sized relationship between increasing uncertainty and decreasing perceived coordination which Guiltinan et al's (1980) was unable to do.The implication of this fmding is that franchisors' efforts to 'absorb uncertainty', or make franchisees feel the environment is more predictable (with regard to sales and profits, for example), are likely to increase perceived coordination; or, vice-versa, if increased coordination is perceived then franchisees may feel less insecure about the unpredictability of the environment.The fmding (of both studies) that there is, indeed, a positive S.-Afr.Tydskr. Bedryfsl. 1984, 15{3) correlation between the helpfulnes of communications (personal and printed) to the franchisee and perceived coordination, is good evidence that franchisors should be concerned with both the quantity and quality of information flows from them to the franchisees.This confirms the importance of area managers, house journals, store visits by franchisor managements, other mechanisms that enhance communication, and a feeling of being part of the team on the part of franchisees.

Table 2
Pair-wise correlation for coordination and authority

Table 3
Pair-wise correlations for coordination and other four areas

Table 4
Comparison of the five areas and their correlations with work coordination