A design without logo does not damage brand : a packaging with a generic logo and without logo vs a symbolic logo

Private label brands have been growing slowly across different countries and have made their place on the shelves and the sales. Some of them have their own different logos from the retailer, others use the same logo and branding. We wonder what would happen to a product without a logo. We explore also what would happen if the distribution logo had a non-symbolic shape rather than a specific figurative new shape, but different from the retail brand. We designed an experiment with 3 groups. We observe that the logo absence and the non-symbolic logo have results on brand perception. The design with a symbolic logo obtains better results. Results suggest that using a symbolic logo, could be interesting for retailers with private label brands.


Introduction
Over the last decades private label brands (PLBs), also called "phantom brands" have increased their magnitude and suppose up to a 31.4% of the assortment in Europe.In the last years, 100% of Spanish homes bought private label products (PLPs) and the PLB market share is around 38.3% (Contreras & Conde, 2018).Even platforms like Amazon have launched or are planning to launch their own private labels.PLPs have had a high impact on the market and consumer perceives the confidence on these products under the umbrella of retailer and not from the producer.
Specifically, in Europe, we observe how some retailers (supermarkets chains) have created PLBs to sell their exclusive private labels while other chains use the same brand as the retail or sales points to mark a uniformed brand (store and product).Other brands offer mixed models, with both new logos and the same logo of the shop or sale point.
In this scenario, we asked ourselves if, under the consumer's eyes, private label brands with no differential sub-brands have lost value and how would a new PLP be perceived if it didn't have a logo.We also wonder if products presented with the same logo as the store or retailer would be perceived differently to a design without a logo.Furthermore, we wonder if it's necessary to segment in different brands or if a single private label retailer brand should be used for all private label products of the same retailer.
New PLB logos can have abstract shapes (the original retailer logo, or not) or can be designed with a logo shape that has a symbolic link with the category product.Parting from this scenario, we planned a basic experimental research focused on these problems.

Subject of Study
Our study subject is the influence of the absence or presence of a logo and its symbolism on the brand perception, on a private label product of low added value.The product is bottled mineral water, the packaging is the water bottle and the labelling is the same for three cases: no logo scenario, a non-figurative (as in abstract) logo and a symbolic logo.
The main research question of this study is how is brand perceived when there is no logo, when there is a logo with a random shape that is non-figurative and when it has a symbolic shape or a figurative shape.

Hypotheses
The main hypothesis is that the presence of the symbolic logo will correlate with a higher brand perception, estimated price, brand preference, brand confidence and purchase intentions.The symbol should have, according to what we will present on the theoretical background, a positive effect on these variables.
Here are the hypotheses per each independent variable: H1: The presence of the brand in form of logo influences positively on declared results of purchase intention, brand preference, price estimation and brand confidence.H2: The packaging design with a non-figurative logo influences positively on declared results of purchase intention, brand preference, price estimation and brand confidence, compared to the absence of logo.H3: The packaging design with a symbolic logo influences positively on declared results of purchase intention, brand preference, price estimation and brand confidence, compared to a design with a non-figurative logo.
We consider the influence of an incremental correlation (positive or negative) observable with

RSCH
specifically in Spain, almost a 40% of the sales.Probably, in 2018, PLBs will constitute a 30% of large-scale consumer market.
PLPs success is motivated by several factors, probably derived from the economic crisis.According to Nielsen (Contreras & Conde, 2018) 81% of Europeans who buy these products do it in order to save money (it's 70% globally and 100% of Spanish families).This is not contradicted by the "excellent price-quality" perceived, because an 11% perceived more quality on PLPs than in manufacturer, producer or traditional brands (Contreras & Conde, 2018).In summary, according to these numbers, consumers seem to accept that the pay for lower price product compromising quality and PLPs are related with low price and low added value.
18% of the new products launched on large scale consumer market in 2012 where PLPs and this percentage went down to a 3.1% in 2016.We could argue that this lack of new PLB products, according to Kantar Panel in 2017, is due to previous launches and that no new products are necessary under private labels (Kantar Distribution and Consumer; 2017).
We carried a brief field study to analyse how retailers use their private label brands and how they make use of their logo or logos.Among PLBs in large scale consumption market we observed 4 different uses (a, b, c and d).Model "a" are private label products dissociated from the retail store.LIDL, or DECATHLON are good examples of this model (figures 1 and 2).These brands have its own logo for the stores, and each category has an exclusive private brand which is not the same as the store identification.The second observed model, or "b", is the one followed by those stores that sell PLPs under the same PLB and matching the store identification (Carrefour, Dia).The same logo is use for PLPs under different categories.Finally, there is the mixed model or "c" of AUCHAN or TESCO.These stores usually has the same store logo and for its PLPs, but they also have differenced PLBs (exclusive of the retailer, as in the other models) for a reduced number of categories (figure 3).
Finally, the a less usual model is "d"; it is the model of ALDI (figure 4), a supermarket chain that does not have own private label brand, nor sells under its own brand or brands.Aldi's assortment does not usually include top selling brands (with a few exceptions like Coca-Cola).

Logos, brands, symbols
There might be some controversy around the statistical tools.Analysis will be based on the null hypothesis test.

Theoretical Background
Private Brands Success and use Private labels are probably a worldwide phenomenon PLPs have grown in the last years and involve a 31.4% of the assortment in Europe, and A design without logo does not damage perception: products with generic logos might have lower brand impressions RSCH term "logo" and it might be misused with the term "brand", "sign".Brand and logo can be considered as synonymous according to their descriptions, although logo is the term more used for the "sign" or "signature" of the brand and "brand" is generally a set of parameters around the identity of a product.
The Oxford Dictionary defines "brand" as "A type of product manufactured by a particular company under a particular name" but also as "A particular identity or image regarded as an asset" (Oxford, 2010).There is some controversy since the second entry can be confused with the logo, being it more specific.The Spanish academy defines the Brand as a "sign that is made or is put in something or someone to distinguish it, denote quality or belonging" (Real Academia Española., 2014).
The American Marketing Association (AMA) defines brand adding a bit of complexity as "Name, term, design, symbol, or any other feature that identifies one seller's good or service as distinct from those of other sellers."(AMA, 2018).
Flórez Calderón defines the Brand as a "material sign fixed on a durable support (Flórez Calderón, 2015) and Frutiger defined brand as "a modern sign" (Frutiger, 1981).
We want to distinguish clearly brand from logotype since the latter can be considered "the graphic representation of a company, institution or organisation's name" "optionally joined with an illustrative element" (Tena Parera, 2005).
Since brands as Nike have a logo without type, we could argue that typography is also optional.The AMA defines "logo" as "A graphic design that is used as a continuing symbol for a company, organization, or brand.It is often in the form of an adaptation of the company name or brand name or used in conjunction with the name" (AMA, 2018).This definition, in our opinion, sets the difference between the brand and the logo as a symbol (with or without type).

Figurative logo or symbolic logo vs non-figurative logo
We observed that PLBs have logos that we classified in 4 marketing models (a, b, c, d).These models can be narrowed in 2 groups regarding their logos: retail matching logos (same brand and same logo for the stores and the PLPs) or a set of independent PLBs in product categores (exclusive of the retailer).We have also observed that the logo is a sign, identified with a brand.
Logos are a graphic design forms by either a letter or a shape or a combination or both (Tena Parera, 2005; Wong, Alsina Thevenet, & Rosell i Miralles, 1995).Leaving out the lettering, these

RSCH
shapes can be considered figurative or non-figurative (Flórez Calderón, 2015;Landa, 2011).A figurative shape has a direct meaning as a symbol that represents a real object (a house pictogram or isotype, a leaf) or they can be abstract shapes (the Nike logo).
Considering that the shape of a new brand or a private label brand can be related or not related to the retail logo, we will find that some PLB logos have shapes with symbols that remind us of the category and other retailers use PLB logos that do not represent anything specific beside the brand that they are linked to, and that this sign will become a symbol when it's widely accepted (Flórez Calderón, 2015).Since the logo is a shape and/or a text, we observe that logos fall in two categories: those that have a symbolic meaning further than the brand (symbolic or figurative) and those that only represent the brand.
In summary, we will consider that a non-figurative logo is a logo (shape and/or type) that does not have an identifiable symbol and that does not represent something figuratively.For instance, the sign of radioactivity is often used as an example of non-figurative and abstract sign (Landa, 2011).We will consider a symbolic or figurative logo that which has a symbology or a figurative meaning because of its shape or the shape of the type.Both types of logos are signs, since they represent the brand.
We consider that this differentiation is very significant specially for new products launched under the retail brand (generic PLBs) and for new products launched as new PLBs by a retailer.

Private Label Brands Logos and Symbolism
Oxylane (the company behind Decathlon stores) owns many brands with their logos for each category (sport type) and no PLP has the Decathlon nor Oxylane brand, only the labels, tickets etc… (Model "a", see figure 1).None of the shown logos can be considered, as per the previous definitions, as symbolic.They are just specific PLBs separated from the retailer brand.
In a similar case (model "a"), figure 2 shows some of the LIDL retailer PLBs, and again, none of them uses the LIDL logo.The BioOrganic brand logo contains a recognizable tree, and so it does Solevita.Both these brand logos are used for PLPs that are related to a tree (vegetables or fruits): the first is used for ecologic products and the second is for fruit juice.Therefore, the tree is a symbol and it is related with each product categories.
Figure 4 shows the mixed model of Auchan.Most products are sold under the Auchan brand logo, with a recognizable bird, and, although it could not have a direct meaning in the mind of the user or consumer, it's still a symbol of a bird.The PLB logo of the Baby brand (used for baby care products) has the shape of a cloud.This shape could be linked to softness, wellbeing and innocence, but not directly linked to a baby, anyway, it is figurative and symbolic, and we can't consider it as non-figurative since it is not abstract.

Methodology
We designed a 3 group experiment for each logo scenario of the study object and the theoretical A design without logo does not damage perception: products with generic logos might have lower brand impressions RSCH background.We created 3 different water bottle designs approved by pre-tests.The 3 models had the same visual identity except for the appearance of the logo and its shape symbology.The experiment was developed with 60 subjects and the expected confidence interval is 95% (p value =0,05).
The test bottle is intended to be a mockup for a new private label mineral water of an undetermined retailer (allegedly with private label brand or brands, but unknown for the subjects).The dependent variables are intended to measure this product brand perceptions on the 3-different logo type groups.

Variables and materials: validation process
The independent variable of this experiment is the logo, with three chosen values: absence of logo, a non-figurative logo or abstract logo and a figurative symbolic logo.The first value is a design that does not contain a logo among its visual elements.The second value is a logo, as seen in model "a" and the Oxylane example, that does not relate with a figure, it is only a sign of this new brand.The third variable value is the presence of a logo that has a figurative meaning besides the signage of the brand (after the pre-tests, the water drop shape).
The dependent variables chosen to represent brand perception are: purchase intention, price estimation, brand preference and brand confidence in Likert or semi-Likert type scales.
Purchase intention is considered as "a decision plan to buy particular product or brand created through a choice/decision process"(AMA, 2018).It was measured with a similar scale adapted from other studies and translated to Spanish/Catalan (Shimp & Sharma, 1987;Taylor, Houlahan, & Gabriel, 1975).Price estimation was scaled in 10 values in ranges of 5cts (euro) from the cheapest to the most expensive similar water bottles in the market (private label and producer brands; from 0,19€ or less to 0,60€ or more) According to a standard definition confidence is "The feeling or belief that one can have faith in or rely on someone or something" (Oxford, 2010).Brand confidence is considered for this study as the confidence felt about the mock up bottle shown to the subjects in each group.
Brand preference was considered as the possibility of this product to be chosen as a substitute David Badajoz, Gemma Freixa RSCH of the current brand consumed, considering that it is not usually a high added value product.The question specified this preference as a "substitution", but measured only by the aspect of it (Colet Ruz, 2013;Martínez Bouza, 2011) and not it's attributes (nourishment and flavour).
Control variables were the packaging or bottle, the label, and its design, overall all bottle elements were homogenised except for the logo presence or type.Subjects were controlled via sociodemographic questions (age, education, province).

Procedures
The process of designing the 3 bottle versions was validated with 60 subjects to evaluate its suitability as the most neutral, plausible and believable designs that would not affect the brand perception except for the independent manipulated variable, the logo.
For the naming, we parted of a brainstorming with 50 names, of which 9 were selected and tested with 60 subjects.As the name would not appear on the label in one experimental group, we chose the less dispersed and most centred result."Fresca" obtained middle results on the suitability question for a water brand, others had too positive, too disperse or too negative results (always contemplating that one label would have no logo since it is our first research objective and question).
A second pre-test evaluated 9 possible figurative and non-figurative shapes as the logo of a water product for a private label brand.The rectangle was selected as the non-figurative shape because it had less dispersion among positive answers, and the water drop was selected as the most adequate among both figurative or symbolic shapes.
On a third pre-test question, the label was designed and printed by the authors (both graphic designers) and it was tested with the 3 values of the independent variable: no logo, non-figurative shaped logo (rectangle with the naming on top of it), and symbolic logo (water drop and the naming aside).The dominant chosen colour was blue, since most brands assorted are using this colour already (as seen on our field study).
On a 5 item Likert type scale, test subjects of this validation test (N=20) answered whether the labels we designed were considered between "very suitable" to "not at all suitable" as a private label water bottle.On both designs with logo the label obtained a 95% of positive response.Without the logo it had only 85% plus a 10% neutral, so, aggregated the neutral value responses were 95% in the 3 designs and very similar regarding their dispersion.
Once validated the 3 labels with the 3 logos, they were printed and glued to standard water bottles, filled with water.The same model of bottle with three different neutral graphic design versions, almost identical differed only by the logo.Subjects 60 participants of the Barcelona area were selected through cultural entities and activities.The Graphic 1: Purchase intention boxplots.Lines show averages.Dispersion is lower and mean is lower in the waterdrop logo (symbolic).
Graphic 2: Brand preference boxplot.Lines show averages.The symbolic logo (waterdrop) has a more condensed distribution, higher mean and lesser dispersion.
A design without logo does not damage perception: products with generic logos might have lower brand impressions RSCH initial intention was to measure brand preference on a retail setting during purchase moment, but a pre-test showed that they were eager to answer and we did not have enough resources and time.The experiment was finally developed on a non-controlled environment, and not a retail set-ting as we would have wished.Participants were reached during leisure time (same hours at the end of the day in weekdays) and they were presented with one of the three bottles, which they could hold and handle during the test.

Results
The statistical analysis used to test the hypotheses was the null test hypotheses with an expected p value of 0.05 (confidence interval of 95%) via Chi squared and correlation direction by Pearson correlation or R value.A descriptive analysis is also presented in order to compare averages and percentages added to the inferential tests and box plots if possible to see the maximum/minimum, averages and quartiles represented graphicaly.
Here below are presented the results for each dependent variable for each experimental group.

Purchase Intention
The purchase intention hypothesis was that the absence of logo would be positive on purchase intentions (on the Likert type scale) and that symbolism of the logo correlates positively on this variable.
Graphic 1 shows on a box plot the higher values obtained by the symbolic logo, as well as similar results for the non-figurative logo and the no logo group.Average was +1.1 on the version without logo, +0.65 for the non-figurative and +1.8 for the purchase of the water drop version.
Signification is 0.000 for the Chi square test which indicates a difference between the influence of the variable with a confidence interval of 99.99%.Pearson's R, confirms a positive statistical correlation (R=0.243) with a 94.9% interval of confidence (Sig.=0.061).These results suggest that the symbolic logo correlated positively with higher purchase intention and results were lower very similar between the design without logo or with the abstract non-figurative version.The average of the non-figurative logo was unexpectedly, the lowest, therefore this is hypothesis is partially accepted for the symbolic logo hypothesis but not the no logo design.

Brand preference
The initial hypothesis was that brand preference would be higher on the group with a symbolic logo and lower without a logo.Descriptive analysis shows a difference between the three groups.The group who evaluated the design without a logo shows an average of 0.25, non-figurative logo is (again) lower at 0.10 and symbolic logo had an average of 0.95 on the Likert scale.
The null hypothesis test shows a significance of 0,007 on the Chi square test, showing that Graphic 3: Price estimation linear distribution representation.Symbolic logo (green) has a higher price estimation among test subjects.

RSCH
they difference is confirmed with a confidence range of 99.93%.Pearson's R results on R=0.227 and a significance of 0.081, and a confidence of 91.9%.Visually, this is shown on the box plot graph (graphic 2) The worse results were obtained (mean and visually) on the non-symbolic logo, and higher on the symbolic logo group.Therefore, the hypothesis is only accepted partially.
For price estimation, the hypothesis was that the design without logo group would express a lower estimated price and a higher estimation on the group shown the symbolic logo of the water drop.The averages show that the hypotheses are confirmed in both cases.The bottle without brand logo was estimated at an average of 3.45 on the scale (equivalent to 0.27€), the non symbolic logo bottle was estimated on average at 4 on the scale (equivalent to 0.30€) and the symbolic logo obtained a 5.4 value on the scale (which is around 0.37€).Chi squared test shows these differences are significant with an α=0.001 (Confidence 99.9%), and confirms a positive 0.475 value and a high significance (α=0.000).Graphic 3 shows a distribution of the scale for the three groups and graphic 4 shows the box plot with the quartiles.

Brand Confidence
The hypothesis is again that the symbolic logo Will result in better brand confidence on the scale, and that the no logo bottle would be negative on confidence.In this case the global average of the three groups was 1.93.The average for the bottle with-out brand logo was a bit higher with 1.95, 1.8 for the non-figurative logo and 2.05 for the symbolic.These results are very close and again worse in the case of the non-figurative logo group.
The Chi squared test shows a significance of 0,01.The Pearson correlation shows a positive influence of R=0.043, but it's significance is 0.743, showing that the confidence is too low (a bit over 25%) to accept the hypothesis.Graphic 5 shows the results of the three groups and the small difference between them.In summary, the homogeneity of the three groups does not allow to accept any of the hypotheses.

Cronbach's Alpha Coefficient between the 4 dependant variables
We tested the reliability of the theoretical dimensions measured via Cronbach's Alpha, which is usually used to validate scales.We wanted to test the opposite, that they are not measuring the same dimension. Results

10
A design without logo does not damage perception: products with generic logos might have lower brand impressions

RSCH
Price Estimation are linked with a α=0.74.The rest of the variables don't get more than 0.55.Since Price Estimation is the only variable that was not measured via Likert type scale, but a range of prices, we can say that they are linked, but clearly not measuring the same dimension.We suggest to research this further in the discussion.
Participants Sociodemographics 100% of the subjects were residents of the Barcelona area, 55% were women and 45% men.25% of participants were 25 year old or younger, 15% up to 30, 15% up to 35 years and distributed in groups of 10-15% from there.55% of the participants were 35 years old or younger.The distribution can be observed on Graphic 6.
Regarding their education, 20% had basic education, 40% had bachelor's degrees and another 40% had university degrees, plus 10% had a master's degree.As for their occupation, 15% were students plus an additional 20% who studied and also worked.40% were workers and 15% were unemployed or retired.

Limitations
We could not develop the experiment on a controlled environment nor a more natural or ecologically valid placement.We could consider it a semi-experiment due to this limitation, although the 3 experiment environments were similar with a randomized test bottles.
The confidence variable had very homogenous results, which could indicate either that it is not of influence or either that the concept or question was not understood by the test subjects.Before discarding it in other experiments, we suggest developing and validating it so we can measure what was intended with more certainty.It was intended to measure that faith or believe that the product's brand would deliver what was expected.
The experiment was limited to the Barcelona province, so it is good for homogeneity of participants, because they had the same cultural and shopping background, but it's also a limitation, so the experiment should be tested in other cultures and countries.
The experiment is limited to new brands.We only tested the effect of the logo absence and its symbolism for a new unknown private label brand.The bad results regarding the non-figurative logo, probably don't apply with existing PLBs and retailers, especially with those that hold a high presence in the mind of the users and consumers.
Finally, regarding the label and packaging design, some participants noticed the blank on the zone intended for the logo.This could have been seen in the lower evaluation during the pre-test of the label design.Anyway, this limitation does not seem to have affected negatively on the evaluations by the overall participants, since the hypothesis of the lack of logo was not confirmed and the worse evaluated design was consistently the non-figurative version.

Conclusions
The results of the design without a logo are surprising since they deny the hypotheses related to it being more negative than any logo (figurative or non-figurative).The hypothesis that the absence of the logo affects negatively the brand perception is not confirmed.The presence of the non-figurative logo was worse on purchase intention and brand preference than the lack of logo.
Regarding the symbolic logo, 3 hypotheses are accepted and one is not, related to confidence, which we would suggest investigating further, validate a scale or test qualitatively.The water drop shape on the logo was positive on price estimation, purchase intention and brand preference, but not confidence.These are the 3 hypotheses accepted.
As for the non-figurative logo, it had worse results than the absence of logo The symbolic logo was not more beneficial for a new hypothetical private label brand.We Graphic 6: Pie distribution of test subjects ages.
David Badajoz, Gemma Freixa RSCH did not expect this result but it suggests that, except for brand confidence, having a random, unshaped or abstract logo, not symbolically related to the category has similar or worse perceptions than the lack of a logo.We suggest that more works are necessary in this line of research with known brands and with a good position in the market.
Regarding the 4 measured variables, three of the four dependent variables were sensible to the changes while confidence was homogenous between the three groups.We also suggest more work with them and following indications regarding "confidence".
Further than the hypotheses that have been confirmed, the implications of this study would apply only to new PLBs on new markets.For instance, observing the results, we recommend that retailers introducing PLBs in new markets use symbols on the logo shapes, related with the product category.The abstract typographic logo did not get good results at all, especially when compared to the absence of logo.
In summary, we suggest that more research about symbolic logos and private label brands should be developed in new categories, with well-established retail brands and compared to similar designs.The non-figurative logo of a well-known retail brand could result in very different responses.
As graphic designers we want to point out that the abstract and quite random logo had these results while the more symbolic, tested and thought logo, and also related to the category (water drop), had much better responses (except for confidence).We know how important is the creative process in graphic design and we insist how the results point in that direction, and how relevant it is for graphic designers to not design randomly.

Figure 1 :
Figure 1: Decathlon sells under different brands, and none of them is Decathlon.

Figure 2 :
Figure 2: LIDL owns many brands with different logos, some of them with symbolism for it's own categories.

Figure 3 :
Figure 3: Mixed model or "c": different PLBs under Auchan group with different logos.

Figure 4 :
Figure 4: Aldi's short assortment is not represented with known brands, but they are not Aldi's brands, it's a selection of products under logos that can or not be related with the product category.

Figure 6 :
Figure 6: The final mockups for the experiment Figure 5: The 9 figurative and non-figurative shapes pre-tested for a water brand

Table 1 :
Purchase Intentions and logo type results.

Table 2 :
Brand preference and logo results.Price Estimation

Table 3 :
Price estimation and logo type results show that only Purchase Intention and