An e-commerce site needs tools that handle five marketing jobs: measuring sales; reaching customers by email, SMS or WhatsApp; moving visitors closer to a purchase while they are on the site; building trust with reviews and social proof; and bringing in new visitors through advertising. For most stores, a sound order is to set up measurement first, then build a list of opted-in customers and switch on automated messages, then add on-site tools and review collection, and grow the ad budget last. A store that reverses this order cannot see why the visitors its ads bring in are not buying, and it has no channel for reaching customers who do not come back.
Once you have finished reading, you will be able to decide which of your current tools are redundant, which jobs no tool is doing yet and where your next budget should go. You will also know which questions to put to a vendor in writing before you buy new software.
What is a marketing tool, and what does it do in a store?
A marketing tool is software that automates or measures one of the jobs a store does to find, persuade and win back customers. Your e-commerce platform’s reporting screen, an email newsletter service, an automation that messages people who leave products in their cart and an ad management dashboard all belong to the same category. What they have in common is that they carry out work you would otherwise do by hand, and they show you the result in numbers.
Online lists of the “best digital marketing tools” put design apps, newsletter services and keyword tools side by side. For a store owner, it is more useful to group tools by the job they do and look at which job is not being done yet. Sort e-commerce marketing tools into five groups this way, and your shopping list gets shorter:
- Measurement and analytics tools show which channel brings in orders and where visitors give up.
- Messaging tools reach customers who have given consent, automatically or in bulk, with email, SMS or WhatsApp marketing messages.
- On-site conversion tools (pop-ups, exit surveys and product recommendations) work while the visitor is on the site.
- Review and social proof tools make the opinions of real buyers, and the interest a product attracts, visible on the page.
- Advertising tools bring in new visitors from search engines and social media, and also reach people who have visited the site before.
What to ask before buying a tool
Dozens of software products do the same job, and their demo screens look alike; when you are choosing among digital marketing tools, that resemblance can easily mislead you. Before you decide, get written answers to these questions:
- Does it have a ready-made integration with the platform you use (Ticimax, ikas, IdeaSoft, T-Soft, Shopify or another platform), or will you have to move order and customer data across by hand?
- Does the results report stop at clicks and opens, or does it also show the orders that follow a message or a pop-up? Within what time frame does an order have to come in for the tool to credit it to that message or pop-up?
- Does it exclude people without consent from your sends, and does it process opt-out requests automatically?
- If you stop using the tool, can you export your customer list and your historical data? Can you take your consent records with you as well (who gave consent, when, and through which channel)?
- Is the price a fixed monthly fee, or does it rise with the number of sends or visitors? On WhatsApp, is the fee Meta charges per template message included in that price, or passed on separately?
- Which country does it keep customer data in, and does it set out the obligations on personal data protection in a written contract with you?
- Is another tool you already have doing the same job?
Do not skip the last question. Your platform’s own campaign module, a newsletter service and a separate pop-up tool can show the same visitor three different discounts on the same day. Before buying a new tool, list the ones you already have and write down in a single sentence the job each one does; if the same sentence appears on two lines, switch one of those tools off. On the day you switch a new tool on, add it to the list together with that date and the number you expect it to change.
Do not add other tools until measurement is in place
Measurement is the first layer to set up, because it is what tells you whether each tool that follows is doing its job. Set aside the first two weeks for this alone and keep to the sequence: launch your cookie consent banner first, and check the figures in analytics only after that. If you run the check before the banner, the figures in analytics may shift once it goes live, and your first check will be out of date.
If you use Google Analytics 4, look at the purchase and refund events defined in Google’s e-commerce measurement documentation. Your platform may already send these events through a ready-made integration; even so, check for yourself that they are being sent correctly. Our article on reading e-commerce customer behavior explains how to spot, in analytics, the step at which visitors drop out on the way to a purchase. Track visitors who arrive from organic search in Google Search Console: Search Console’s Performance report shows which queries bring traffic to your site, how many times your site appeared in search results and how many times it was clicked.
Take last week’s order count from your platform’s admin panel and compare it, day by day, with the number of purchases in analytics. If the gap is concentrated on particular days, look at the campaign or payment method on those days. The gap can come from orders paid by bank transfer, from cancellations and returns, or from visitors who reject the analytics cookie or use an ad blocker; note down once how each of these shows up in analytics. Leave a gap unexplained and you will also misread the sales that an ad or a message flow brings in.
The second part of measurement is customer data. Knowing which customers have not bought anything in the last six months, or which ones buy from a single category and never from the others, determines how well every message you send later hits its target. Customer segmentation therefore counts as a continuation of the measurement layer. Pushouse’s Website AI Reporting feature turns a sentence such as “customers who bought trousers in the last 60 days” into a customer group based on purchase history. A group made up of customers without communication consent, however, can appear on the card but cannot be turned into a messaging list.
Cookie consent is part of the measurement plan
When you add analytics and ad tags to the site, plan cookie consent at the same time. Türkiye’s Personal Data Protection Authority (KVKK) states in its Guide on Cookie Practices that cookies used for behavioral advertising require explicit consent. The guide takes a more flexible approach to first-party analytics cookies: where these cookies are used only to produce anonymous statistics and do not track a person’s browsing across different sites, it considers that they can rely on processing conditions other than explicit consent. Among the measures a data controller can take, the guide lists not sharing the collected data with third parties; it also states that if data is transferred abroad through cookies, Article 9 of Law No. 6698 on the Protection of Personal Data must be complied with.
If you use an analytics tool that sends its data to another company, the more cautious choice is to place it alongside the tools that wait for the visitor’s consent. Decide with your legal adviser which tag falls into which group, then record each visitor’s choice with a cookie management tool. We cover the wording of the cookie notice and the consent process separately, in our article on the cookie policy.
Build email, SMS and WhatsApp tools on consent
Once measurement has settled in, give weeks three and four to the messaging channel, and start by adding a separate messaging consent checkbox to the sign-up and checkout steps. Messaging tools let you reach customers who have given consent without entering an ad auction; even so, you do not set all of the channel’s rules yourself. Meta’s WhatsApp pricing documentation states that the charge applies per delivered template message, and that Meta can update its prices at most four times a year, at the start of a quarter. On top of that, Meta’s documentation on per-user limits explains that WhatsApp may limit the marketing template messages one person receives from any business within a given period; a message that hits the limit is not delivered to that person.
Rather than sending the same announcement through all three channels at once, give each channel a job of its own. Email suits long-form content and newsletters, as well as documents meant to be kept, such as invoices. SMS works for one-line notifications. WhatsApp, where customers can reply and messages can carry images and links, comes into its own for personalized messages such as cart reminders, order and shipping notifications, and back-in-stock announcements.
Whichever channel you choose, review your consent process before connecting the tool. Under Law No. 6563 on the Regulation of Electronic Commerce, electronic messages sent for commercial purposes (commercial electronic messages, in the law’s own terms) require the recipient’s prior consent, and recipients can refuse such messages at any time without giving a reason.
Also register the consents you collect in İYS (İleti Yönetim Sistemi), Türkiye’s Message Management System. The Ministry of Trade’s İYS page describes the system as the central structure where the message consents given by citizens are collected. Under the regulation amendment published in the Official Gazette (Resmî Gazete) on January 4, 2020, a business that wants to send commercial messages must register with İYS, and commercial messages cannot be sent to anyone whose consent is not registered in İYS. The amendment also requires consents you collect outside İYS to be recorded there within three business days, and any consent not recorded in İYS is deemed invalid.
The WhatsApp Business Messaging Policy, too, allows a business to message only people who have given it their contact details and opted in to receive subsequent messages from it. The policy also requires you to respect requests to stop messages or to opt out, whether they are made on WhatsApp or elsewhere. A customer who wants to opt out may tell you so by email or by phone, so set up a way to record those requests in your list as well. The first message a customer receives after opting in at sign-up or checkout should say clearly what they have joined and how to leave:
Hi {{customer name}}, you’re now signed up for campaign and new product announcements from {{store name}}. We’ll send you a few messages a month. If you’d rather not receive them, just tap “Stop promotions” below.
Automated flows first, bulk campaigns second
Once the messaging tool is set up, the first job is to switch on the flows that a customer’s own action triggers. A reminder to someone who left their cart, order status notifications and a back-in-stock announcement give customers useful information at exactly the moment they need it. In a cart reminder tool, first check whether customers who have already completed their order are being sent a “you forgot your cart” message. Our WhatsApp cart recovery module does not send that message to a customer who has placed their order, and it does not send a second message for the same cart. For how to adjust the delay to the product and the cart value, see our article on the timing of cart messages.
Bulk campaign messages are the second stage. As the list grows and customer groups become clearer, you can send different offers to different groups instead of the same text to everyone; our article on customer groups for campaign messages explains which groups to separate. When choosing a bulk sending tool, check whether its report shows the orders placed after each send.
For on-site conversion tools, find the obstacle first
In the second month, it is time for the tools that work while the visitor is on the site; some of them also grow your messaging list. Collecting contact details through a pop-up, asking a short question at the moment of exit and suggesting complementary items on a product page are typical examples of this group. The share of people who put a product in their cart and leave without buying is too high to dismiss: the average cart abandonment rate that Baymard Institute calculated from the data of 50 studies is 70.22%.
Which tool to set up first depends on the step where you lose visitors. For visitors who view a product page and leave without adding to cart, look first at what information or trust is missing; product recommendations, social proof and reviews work at this step. For visitors who add to cart but stop short of checkout, turn to the exit survey. From a first-time visitor who leaves without touching a single product, you can ask for contact consent rather than an order at this stage; lead capture pop-ups are set up with this visitor in mind.
With an exit-intent popup, you can ask a visitor who is about to leave the site, “What stopped you from completing your purchase?” The answers, given through options such as price, delivery and product information, show you which page to fix first. Our article on the e-commerce exit survey covers which pages to show the survey on.
Lead capture pop-ups, used to grow the contact list, work when they tell visitors plainly why they should join. Choose a single promise, such as news of a new collection, a campaign announcement or a welcome discount, and if there is a discount, state its scope and expiry date next to the form. On the form, ask for messaging consent with a checkbox that is separate from the discount offer. For most stores, a reasonable rule to start with is never to open two pop-ups in the same page view and to show no pop-up at all at checkout.
Product recommendations based on sales data, shown on the product page, also belong in this group. The link between the recommendation and the main product should be obvious to the visitor, as when someone looking at a phone case is shown a screen protector for the same model.
Show only verifiable data in reviews and social proof
Reviews, and notifications along the lines of “how many people viewed this product today,” give a visitor who has not seen the product before something to base a decision on. You need to be able to prove that every review you show came from a real buyer, so choose your review tool by that criterion. The Ministry of Trade’s announcement on Advertising Board decisions notes that businesses have been given a verification obligation to prevent fake, misleading or unverifiable reviews. Under that obligation, reviews and ratings must come from consumers who bought the product or service in question.
A practical way to ask for reviews only from people you know are buyers is to send the request, linked to the order record, to the customer whose order has been delivered. Since the review request also goes out through your messaging channel, you can leave this step for the third month, once the channel is up and running.
Our WhatsApp feedback module sends the customer a short survey on WhatsApp after an order is delivered; a review you approve appears on the product page with a verified badge, because it is matched to the order record. Do not weed out negative reviews at the approval step. Under the amendment to the Regulation on Commercial Advertising and Unfair Commercial Practices that came into force on August 1, 2026, reviews must stay published for at least one year after the necessary checks, without any distinction between positive and negative ones, and when you do not allow a review to be published, the customer who wrote it must be told right away.
Apply the same standard to social proof notifications: the number of views or purchases in a social proof message must match the real data in your store. A customer who sees a “last 3 items” warning that does not match the actual stock, and finds it still there the next day, may start to look at your price and delivery information with suspicion as well.
Grow the ad budget once measurement and messaging are ready
Search and social media ads are among the digital marketing tools that bring visitors to a store fastest; at the same time, because you pay for every impression or click, any measurement error hits your budget directly. Increase the budget step by step, after you have seen that the measurement adds up and the message flows are working. By the Baymard average above, even among people who add a product to their cart, most leave without buying. A store that does not collect the contact details of visitors who arrive through ads, and does not send a reminder to people who leave their carts, ends up paying for ads every month to try to reach the same person again.
According to the KVKK guide mentioned above, behavioral advertising cookies used for retargeting require explicit consent. Set up your ad tags to run only for visitors who have given consent, and plan your budget knowing that your retargeting audience will stay smaller than your total number of visitors. The regulation amendment we mentioned in the section on reviews also sets rules for targeted advertising. For ads shown by analyzing a consumer’s online behavior or personal data, the consumer must be told, directly and in an easily accessible way, by which criteria the ad is shown and how those criteria can be changed. Clarify with your legal adviser whether that information will be given on the ad platform or on your own site.
How far you can trust the results in your ad dashboard depends on the purchase data that flows from your site to the ad platform. Meta’s Conversions API documentation says the interface is designed to create a connection between Meta’s systems and the marketing data in an advertiser’s server, website platform, mobile app or CRM software. The documentation also asks you to check, after setup, that events are deduplicated and matched correctly. That check matters even more if you send the same purchase both from the Meta Pixel in the browser and from your server. Meta’s deduplication documentation states that such setups must put a deduplication method in place so that duplicate events can be identified; in the recommended method, both sources send the same event name and the same event ID (event_id). Without deduplication, the same purchase can be counted twice and the ad’s return looks higher than it really is.
If you are starting with a small budget, a campaign limited to a few of your best-selling products is easier to track than a broad-audience campaign. To show your products in Google’s Shopping ads and free listings, you upload your product data to Google Merchant Center; Türkiye is on Merchant Center’s list of supported countries. If you advertise on searches for your brand name, allow for the fact that some of those orders could have come in without the ad, and do not judge that campaign’s return by the same yardstick as the others.
Measure a campaign’s success by the orders you see in analytics and in your platform’s admin panel; the ad dashboard counts the same sale by its own rules. Tagging your ad links with UTM parameters makes it easier to tell in analytics which campaign the visitors who placed orders came from. Send us the list of tools you use and the name of your platform through the contact form, and we can work out together which jobs are still not covered.
